Skip to main content

Full text of "Journal of the Society of Motion Picture Engineers"

See other formats

From the collection of the 

n m 


. a 


San Francisco, California 


Vol 49 JULY 1947 No. 1 



Report of the President LOREN L. RYDER 1 

A Combination Scoring, Rerecording, and Preview 


An Analysis of Low-Reflection Coatings as Applied to 
Glass W. P. STRICKLAND. 27 

Recent Developments of Super-High-Intensity Carbon- 
Arc Lamps . M. A. HANKINS 37 

A Newly Developed Light Modulator for Sound Re- 
cording G. L. DIMMICK 48 

The Physical Properties and the Practical Application 
of the Zommar Lens FRANK G. BACK 57 

Special Cameras and Flash Lamps for High-Speed 
Underwater Photography ROBERT T. KNAPP 64 

Photographing Things to Come M. W. WARREN 82 

A Stabilization System by Rate Measurement 


SMPE Stan 7 Changes 93 

Copyrighted, 1947, by the Society of Motion Picture Engineers, Inc. Permission to republish 
material from the JOURNAL, must be obtained in writing from the General Office of the Society. 
The Society is not responsible for statements of authors or contributors. 

Indexes to the semiannual volumes of the JOURNAL are published in the June and December 
issues. The contents are also indexed in the Industrial Arts Index available in public libraries. 









A. C. HARD* 


**President: LOREN L. RYDER, 

5451 Marathon St., Hollywood 38. 
** Past-President: DONALD E. HYNDMAN, 

342 Madison Ave., New York 17. 
**Executive Vice-President: EARL I. SPONABLE, 

460 West 54th St., New York 19. 
^Engineering Vice-President: JOHN A. MAURER, 

37-0131st St., Long Island City 1, N. Y. 
** Editorial Vice-President: CLYDE R. KEITH, 

233 Broadway, New York 7. 
^Financial V ice-President: M. RICHARD BOYER, 

E. I. du Pont de Nemours & Co., Parlin, N. J. 
** Convention Vice-President: WILLIAM C. KUNZMANN, 

Box 6087, Cleveland 1, Ohio. 
** Secretary: G. T. LORANCE, 

63 Bedford Rd., Pleasantville, N. Y. 
* Treasurer: E. A. BERTRAM, 

850 Tenth Ave., New York 19. 

**JOHN W. BOYLE, 1207 N. Mansfield Ave., Hollywood 38. 

*FRANK E. CARLSON, Nela Park, Cleveland 12, Ohio. 

*ALAN W. COOK, Binghamton, N. Y. 
**ROBERT M. CORBIN, 343 State St., Rochester 4, N. Y. 
**CHARLES R. DAILY, 5451 Marathon St., Hollywood 38. 
"tjAMES FRANK, JR., 356 West 44th St., New York 18. 

*JOHN G. FRAYNE, 6601 Romaine St., Hollywood 38. 
**DAVID B. JOY, 30 East 42d St., New York 17. 

*PAUL J. LARSEN, 1401 Sheridan St., Washington 11, D. C. 

*WESLEY C. MILLER, MGM, Culver City, Calif. 
**HOLLIS W. MOYSE, 6656 Santa Monica Blvd., Hollywood. 
*JA. SHAPIRO, 2835 N. Western Ave., Chicago 18, 111. 
*WALLACE V. WOLFE, 1016 N. Sycamore St., Hollywood. 

"Term expires December 31, 1947. {Chairman, Atlantic Coast Section. 
**Term expires December 31, 1948. ^Chairman, Midwest Section. 
"Chairman, Pacific Coast Section. 

Subscription to nonmembers, $10.00 per annum; to members, $6.25 per annum, included 
their annual membership dues; single copies, $1.25. Order from the Society at address abov 
A discount of ten per cent is allowed to accredited agencies on orders for subscriptions 

and single copies. 
Published monthly at Easton, Pa., by the Society of Motion Picture Engineers, Inc. 

Publication Office, 20th & Northampton Sts., Easton, Pa. 

General and Editorial Office, Hotel Pennsylvania, New York 1, N. Y. 

Entered as second-class matter January 15, 1930, at the Post Office at Easton, Pa. 

under the -Act of March 3, 1879. 


Vol 49 JULY 1947 No. 1 


This is the semiannual report of the president to the members of the 
Society of Motion Picture Engineers. It is a statement of activities 
subsequent to the report of D. E. Hyndman, past-president, and John 
A. Maurer, engineering vice-president, titled "Past and Future 
Society Activities" as published in the September, 1946, issue of the 
JOURNAL. This report was compiled as of Apr. 15, 1947. 

The Soc'ety membership has now reached an all-time high of 2537 
members. The cash and the negotiable security assets of the Society 
stand at $103,793.36 as of March 31, 1947. Your past-president, 
D. E. Hyndman, who is in charge of gaining sustaining memberships 
has reported $14,850.00 received thus far in 1947. It is hopefully 
expected that this amount will be increased by several thousands of 
dollars. The present successful status of the Society places upon the 
Officers and Board a well-appreciated responsibility of continuing 
the advancement and success of this great work. 

As president of the Society of Motion Picture Engineers, I am very 
happy with the way that the Officers and Committees are functioning. 
The Convention now in session is indicative of the increasing activity 
and increasing scope of Society work. In our past experience a good 
normal papers' program has included about fifty to fifty-five papers. 
For this Convention there were over seventy-five papers offered. 
We are also enjoying one of the largest Convention registrations and 

In carrying forward the work of the Society, the Board for clarifica- 
tion has redefined the scope of our work. The Society of Motion 
Picture Engineers is interested in and will participate in all technical 
phases of pictorial rendition of action, whether it be from film as in 
motion pictures, electronics as in television, or other device. The 

* Presented Apr. 21, 1947, at the opening session of the 61st Semiannual Con- 
vention of the Society in Chicago. 


2 RYDER Vol 49, No. l 

Society and its engineers have in the published works of the JOURNAL 
most of the information now required in the art of television, for pic- 
torial recording, trick photography, projection, color, editing, test 
films, lighting, and studio techniques. This and all other informa- 
tion of the Society are to be applied to this now-expanding art. The 
Society is not now and does not contemplate overlapping in the ac- 
tivities of other societies in the fields of radio and radio transmission. 

As most of you know, Paul J. Larsen and representatives of the 
Society of Motion Picture Engineers have appeared before the Fed- 
eral Communications Commission and during 1945 and 1946 ob- 
tained for the motion picture industry frequency allocations for 
theater television use. On petition of other interests the FCC 
issued Public Notice No. 97615 on Oct. 22, 1946, calling for a rehearing 
and reallocation of these frequencies to the exclusion of theater tele- 
vision. After a discussion with Eric Johnston, of the Motion Picture 
Association, Byron Price, of the Association of Motion Picture Pro- 
ducers, Inc., Donald Nelson, of the Society of Independent Motion 
Picture Producers, Inc., and Y. Frank Freeman, chairman of the Re- 
search Council, the SMPE submitted a brief and Mr. Larsen again 
appeared before the FCC at its hearing on Feb. 4 of this year. 
The Society received telegraphic support from both Mr. Johnston 
and Mr. Nelson. 

It is the hope and belief of those familiar with these deliberations 
that action favorable to the motion picture industry will be handed 
down. It is also the opinion of those close to this work that the right 
of this industry to these frequencies will again be challenged unless 
this industry expresses a sincere interest in theater television. The 
Society is therefore calling this subject to the attention of top repre- 
sentatives of the motion picture industry and asking that they as- 
sume future responsibility in this regard. 

The Society is carrying forward the standardization program pre- 
viously undertaken. This includes conversion of war standards into 
American Standards and a new activity whereby the Society, as a 
member group of the American Standards Association, is to co- 
operate in the activities of the new International Standards Organiza- 
tion recently established by the United Nations Organization. 

The Society and the Research Council of the Academy have es- 
tablished a co-operative program for the production and sale of test 
films. In the future all picture and sound test films for 35-mm and 
16-mm equipment are to be released under a joint SMPE and 


Research Council banner. These test films may be purchased from 
either the Research Council or the SMPE. This move has been 
made in order to avoid duplication of effort and to serve the indus- 
try better. 

By action of the Board of the Society, the 62nd Semiannual Con- 
vention which is to be held at the Hotel Pennsylvania, New York, 
Oct. 20 to 24, inclusive, will emphasize theater engineering. 
There will .also be a comprehensive exhibit and demonstration of 
theater material and equipment having new features of engineering 

Respectfully submitted, 
LOREN L. RYDER, President 




Summary This paper discusses the construction of the new Republic Productions 
scoring stage and includes a description of the electrical equipment used for the recording 
of music. In the building of the stage, which is probably the largest scoring studio in 
the world, aesthetic elements were given equal consideration with acoustic factors. 

It was also deemed important that the enclosure contain all of the necessary facilities 
for music recording, such as a dual-reverberation chamber with a remotely controlled 
door, a vocal room with a large window between it and the stage, two monitoring rooms 
with concealed speakers, a conductor's podium, and an efficient air-conditioning system. 

During the past three years, increased production schedules at Re- 
public Studios demanded additional music recording facilities. Since 
the original scoring stage, built in 1938, was a combination dubbing and 
scoring room with compromises in acoustics, it was decided to provide 
an entirely new and, if possible, an ideal music recording studio. The 
importance of such an enclosure had always been minimized in the in- 
dustry, resulting in either multipurpose or converted production 
sound stages. Because of this attitude, no music recording stage of 
appreciable size had been designed and built with the primary object 

* Presented Oct. 21, 1946, at the SMPE Convention in Hollywood. 
** Republic Studios, North Hollywood, Calif. 
t RCA Victor Division. Radio Corporation of America, Hollywood. 


of a finished scoring stage, combining all associated requirements. It 
was therefore intended that the new stage should overcome the short- 
comings of what had gone before and should, in addition, provide 
new and original conveniences for producing better, that. is, more 
efficient and more natural, music recordings. 

In keeping with the modern trend in architecture, to combine the 
aesthetic with the functional, careful consideration was given to the 
appearance of the stage as a whole. It was felt that it would not be 
sufficient to provide just a space for the accommodation of an orchestra 
since this was what all the other stages had done. If the musicians 
were to be treated as artists, which certainly they were, it was essen- 
tial to provide an environment which, instead of being unsightly and 
depressing, should be attractive and inspiring. Certainly, if the boiler 



fH 2.o, 



Number of Instruments 
FIG. 1. 

room of an aircraft plant could be painted in an attractive color 
scheme the scoring stage of a motion picture studio could be made 
clean-cut and appealing. This is no idle repetition of an unimpor- 
tant trifle, but a serious report which considered not merely the 
physical, but also the equally important psychological factors which 
go- into the making of a sound track. 

When the scoring of music is the primary purpose of such a room 
(as it was in this instance), it is essential to determine first the number 
of musicians which such a room is, on the average, to accommodate. 
Fig. 1 gives the number of cubic feet per instrument which, according 

July 1947 


to experience, a scoring stage should have to provide satisfactory re- 
sults. If this allotted space is less than indicated on the figure or 
as on extreme case, if a large orchestra is placed in a small room the 
lack of comparatively long-time reflections becomes easily evident. 
Music tends to lose definition (because of the excess number of short- 
time reflections) and is then spoken of as "blurred" or "cramped". 
Placing a small band in a large room is not nearly so objectionable, 
because a few "flats" or sound-reinforcing panels stationed around the 
musicians quickly lend to the music the desired character. 

For a mean capacity of 70 musicians, the cubical content of the 
scoring stage therefore comes close to 210,000 cu ft. Having deter- 
mined the volume, the matter of deciding on the dimensions of the 
room came next in attention. Since the length of the stage was fixed loo,ooe 5*0,00* 

Volume (Cu Ft) 
FIG. 2. Reverberation time versus volume. 

by the available space (112.5 ft), and since it was desired to make the 
stage part of a three-story building, the height and width finally selected 
for the room were 32 ft and 72 ft, respectively. This permitted the 
construction of a vocal room and a reverberation chamber as adjuncts 
to the orchestra shell, thereby providing the stage with the desired 
210,000 cu ft of volume. The mean height, width, and length of the 
stage (on account of the configuration of the confines of the stage) 
are 29 ft, 64.5 ft, and 112.5 ft, respectively, which conform closely 
with published data on the subject. It should be emphasized, 


however, that blind adherence to "optimal-ratio" recommendations 
may give rise to considerable difficulties in the product. The most 
important consideration in regard to the proportions of a room must 
be the purpose or purposes to which it is to be put. The reverbera- 
tion time at 1000 cycles for a recording studio of 210,000-cu-ft volume 
should be 1.15 sec (see Fig. 2); hence the mean absorption of the 
room (the interior surface of which is 24,000 sq ft) can be determined 
by the equation 

0.049 V 
-log e (1 - a) = TS 

= 0.049 X 210,000 
1.15 X 24,000 

or a = 0.31 

where a = mean absorptivity 

= A/S 

A = total absorption 

3 = total interior surface 

V = volume 

T = reverberation period. 

Hence the total absorption can be computed to be 

A = aS 

= 0.31 X 24,000 . 

= 7440 sabines. 

Next in importance to the total absorption required in the room at 
1000 cycles come the questions of the distribution and the type of the 
sound-absorbent and reflective materials within the enclosure. Here 
we have to consider some factors less easily circumscribed by mathe- 
matical notation. One of these deals with the allocation of the 
acoustic and reflective materials within the "shell". In the early 
days of sound recording (and broadcasting as well), the "live-end 
dead-end" studio was almost universally employed. Such a room 
leaves much to be desired, however. Admittedly, the band is situated 
in surroundings of considerable localized reverberation, thus providing 
the type of environs much appreciated by musicians. However, mul- 
tichannel recording setups began to appear, by means of which a 
sound track may be made for each of a group of instrument sec- 
tions. Also, multiple-microphone pickup conditions proved desirable, 
whereby a mixer blends tl^e individual sections for the purpose of 
achieving a particular "blend" or "balance". For these two cases 
the live-end dead-end studio was found inadequate. 


At first consideration it would appear that one microphone would 
facilitate the mixer's work, in that a considerable tone fusion, existing 
at some distance from the instruments, would ensure instrument bal- 
ance. However, this is not necessarily true since a microphone is a 
monaural device without the ability to select and reject sounds as the 
human ear does when a listener is concentrating. Therefore, in order 
to maintain the definition experienced by a person listening with both 
ears to an orchestra, the single microphone would have to be placed 
closer to the source of sound. With a short microphone distance the 
element of balance between the different instrument sections becomes 
quite critical. The use of multiple microphones, however, obviates 
the necessity for such critical placement, but selective pickup of a 
more or less evenly spread-out orchestra can be obtained only if the 
shell itself is not too live. Indeed, the more microphones are used, 
the deader should be the shell. If it is then desired to add a rever- 
beratory character to the recording, this can easily be done in rere- 
cording by making use of the reverberation chamber with its adjust- 
able reverberation characteristic. 

It must also be remembered that the ratio of reflected-to-direct 
sound at the microphone controllable by changing the microphone 
distance has a pronounced influence in providing an impression of 
reverberatoriness (there is no other word for this sensation, ' 'rever- 
beration" being not sufficiently expressive of the continuous sound- 
reinforcement during transient signals). Hence, if an instrument sec- 
tion as heard through the microphone requires more of this quality, 
this can be secured easily by increasing the distance. The nicety of 
this adjustment is considerably greater than that which can be achieved 
by controlling the acoustics of certain sections of the shell. The 
chief advantage of such a "subdued" shell lies therefore in the fact 
that there is less acoustic spillover from one instrument section to 
another. It should not be thought, however, that this shell is ' 'dead" ; 
far from it; its absorbent and reflective wall sections are approxi- 
mately equal. On the other hand, the so-called "dead' ' end is no longer 
so absorbent, but exhibits a comparatively large area of reflective con- 
vex splays. The round-the-room reflections so often discussed in 
the literature and so little heard in practice are thereby noticeably 
increased, as befits the music of a large band usually heard in a 
spacious auditorium which by conventional construction favors such 
long-time reflections. 

There is another reason why the live-end dead-end studio should 



be avoided. Acoustically speaking, such a studio is no longer a room, 
but merely three walls, a ceiling, and a floor, because reflections from 
the fourth wall (the dead-end) are noticeably absent. Such a condi- 
tion gives rise to an irregular decay characteristic which does not lend 
a pleasing character to the recorded music. Fig. 3 shows the rever- 
beration characteristic of the stage as measured with a high-speed 
level recorder. Next in importance come the shape and the material 
for the room. Shape and material cannot be divorced from each 
other, because the material frequently influences (and in no minor 
manner) the shape or contour of the walls and ceiling. For instance, 
if convex reflective splays (so-called polycylindrical diffusers) are em- 
ployed as part of the interior treatment, it makes a considerable 
difference in the required number of such splays whether they are made 
of plaster or of plywood. If made of plaster, fewer will be required 
than when made of plywood, because, with practical thicknesses of 
plywood and plaster, the plywood splays will be more absorptive at 
practically all audio frequencies than the plaster splays. 

60 60 100 

00 300 400 600 1.000 tOOO TjOOO 4/JOO 5/XX) 10,000 

FIG. 3. Reverberation characteristic of Republic Studio scoring stage No. 12. 

Regarding the shape and the configurational details of the room, 
it was thought, early in the period of planning, that cylindrical wood 
splays would prove desirable for the band-shell contours as well as for 
the side wall and ceiling of the stage outside the shell. Their plenti- 
ful use in another, smaller stage had given ample evidence of their 
effectiveness their pleasing tonal response, their high dispersion ca- 
pacity, and their great absorbent qualities at the low frequencies. 

Fig. 4 shows the plan of the stage, and Fig. 5 its elevation. A 
number of salient features at once will be evident. The shell, of 
trapezoidal shape, and of a depth amounting to practically half the 
length of the stage, affords easy accommodation of orchestras ranging 


from 20 to 100 instruments. The permanent three-riser platform 
extending in an arc across the entire width of the shell represents a de- 
sirable construction in a stage of this size. It not only simplifies 
greatly the arrangement of an orchestra, but provides easily ac- 
cessible outlets (at the steps) for microphones, headphones, and 
music stand lights and thus reduces hazardous cables laid across the 
floor. The shell is confined by painted splays and panels of absorb- 
ent material laid with smooth joints and rough outside surface. It is 

FIG. 4. Floor plan of stage. 


FIG. 5. Elevation of stage. 


lighted without glare by reflector-type, flush-mounted fixtures with 
diffusing lenses, and altogether conveys a pleasing impression with- 
out appearing overluxurious. 

Adjoining the stage are a reverberation chamber and a vocal room. 
A reverberation chamber provides a very necessary adjunct for sound- 
on-film recording, the occasions being indeed numerous when a rever- 
beratory quality is to be added to a recording during or after its com- 
pletion. Briefly described, the process consists in reproducing sound 
in a highly reverberant room the so-called reverberation chamber 
and mixing the output from a microphone in this room with the origi- 
nal recording by a method known as "dubbing" or "rerecording". 
Surprisingly, when the electric level of the "reverberated" signal is as 
much as 20 db below the electric level of the original recording at the 
mixing panel, the combined reproduced signal conveys a strong impres- 
sion of reverberatoriness in every syllable of speech or passage of music. 

Unlike other electric or mechanical means of adding a reverbera- 
tory note to a recording, the chamber method provides both the 
proper growth characteristic and the decay quality of sound in a live 
enclosure. Delay networks, magnetic-tape recordings, and other 
devices for achieving synthetic reverberation usually permit only 
the provision of the decay characteristic; no attempt is made to intro- 
duce the growth characteristic. 

By dividing the chamber into a small and a large room, two differ- 
ent characters of the reverberated signal may be had by placing the 
microphone in either one or the other of the enclosures (assuming 
each has been equipped with a loudspeaker for reproducing the signal) . 
By proportioning the dimensions of the rooms carefully, the spectral 
distribution of the normal modes of vibration can be made decidedly 
differerent in the two rooms and with it, of course, the quality of the 
signal. If a door is provided in the partition between the two 
chambers, additional signal qualities can be obtained by reproducing 
the signal in the room opposite that in which the microphone is located. 
By adjusting the opening of the door, the character of the sound may 
gradually be varied over a considerable range, since the sound-trans- 
mission characteristic of the aperture is a function of the size of the 
opening. The smaller the opening, the smaller the amount of low- 
frequency sound which can be transmitted. The door in this respect 
acts in the nature of a high-pass filter. In the dual-reverberation 
chamber pictured on the plan, the aperture of the door between the 
two rooms was made controllable from the mixing console on the stage. 


The reverberation chambers were built of concrete, and the insides 
of the rooms were finished with cement plaster to secure as high a re- 
flectivity as possible. The walls, as well as floor and ceiling, were 
kept nonparallel to avoid flutter echoes in the rooms. 

The vocal room likewise represents a necessary facility for the re- 
cording of music with vocal renditions. The orchestration and the 
song, in many of such instances, are recorded on separate tracks 
chiefly to have some control over the desired "mix" of the two after 
the respective scene has been photographed, since production require- 
ment cannot always be anticipated fully. Then too, if necessary, 
either the music or vocal number can be replaced later without trouble 
because sufficient acoustic isolation exists between stage and vocal 
room to obtain essentially pure music and vocal tracks. In practice, 
however, a third sound track is always recorded simultaneously with 
the other two, which is a "mix" of music and song as deemed best at 
the time and is frequently the one used in the completed picture. 

Such an arrangement requires that the conductor in the scoring 
stage shall be able to listen to the volcalist as well as to the orchestra. 
He will, therefore, be required to wear one earphone connected to the 
vocal recording channel, while the singer may either wear earphones 
connected to the music recording channel, or else listen to the music 
reproduced in low tones over a public-address system installed in the 
vocal booth. 

The director's podium is mounted on rubber-tired casters with a 
jack- type brake to prevent its moving when in use. The podium is 
provided with a public-address system connected to the mixer and to 
the vocal room, and with headphone outlets with volume control. 
Indirect lighting is provided from a hooded fixture across the entire 
width of the podium, in addition to speed lights and an electric stop 

During postscoring with a picture on the screen for cuing, the 
stage lighting is usually off; to enable the musicians to see the director, 
a spotlight in the ceiling is directed at the podium, resulting in ample 
light on the director without causing any glare. 

Two rooms, 7 ft by 30 ft each, back of the screen and across the entire 
width of the stage, provide storage space for chairs, music stands, and 
instruments. A 7- by 5-ft room located at one end of the vocal room 
is equipped with cabinets for microphones and headphones, and cable 
hooks for all cables when not in use on the stage. 

The scoring monitor room is accessible from the stage proper and 


has good visibility over the entire stage and vocal room. It was felt 
that a large room was desirable to preclude the necessity of low moni- 
tor volume levels, which usually results in loss of perspective and 
good orchestral balance. The monitor speakers (low-frequency and 
high-frequency horns) are mounted flush in the wall opposite and di- 
rected toward the mixer. The mixer console is placed adjacent to the 
viewing window of triple plate glass set at dissimilar angles. The in- 
terior finish of the monitor room is comparable to that of the stage. 
Polycylindrical splays are provided in sequence on one side of the en- 
closure, with the remainder of the walls finished in acoustic tile and 
acoustic plaster. 

FIG. 6. New machine room. 

The original installation 1 of the fixed recording equipment consisted 
of two complete recording channels. One was normally used for dub- 
bing and the other for scoring, although both could be employed for 
two-channel scoring sessions. Eight soundheads, two film recorders, 
an acetate recorder, and the amplifier racks, located in one room, com- 
prised the installation. 

It was thought that more equipment was necessary, with certain re- 
finements that would provide greater flexibility. The original eight 
soundheads, therefore, were moved to an adjoining room and eight new 
heads were added (Fig. 6). Fig. 7 shows the floor plan of the film 



machine room. Sixteen film reproducers are located in the room with 
five machines along each side wall and three pairs of machines, back 
to back, in the center of the room. Three gutters mounted in the 
walls behind the machines carry the speech lines, the exciter lamp 
supply lines, and the interlock and rewind power lines. Loop racks 
are installed over the machines located along the walls. These pro- 
vide ten reproducers with loop facilities. Each machine has a com- 
bined feed mechanism and electric rewind magazine as shown in Fig. 8. 

FIG. 7. Plan of new dubbing-machine room. 

Since all original film recording is Class B push-pull, 2 the heads 
are aligned for a very accurate azimuth and push-pull balance. 

Available space in the original room was utilized for seven more 
amplifier bays, making fourteen in all; three film recorders, and one 
fixed acetate recorder, making a total of five film recorders, and two 
fixed acetates. With four complete recording and monitor channels 



FIG. 8. Film reproducer. 

available, it is possible to re- 
cord a musical on four separate 
channels. With the use of 
sound trucks, additional chan- 
nels can be obtained. This 
possibility was given con- 
sideration in the design of the 
mixer consoles for the monitor 
room and the review room. 
It was thought that if the new 
scoring-mixer console were 
engineered for a split channel, 
the review room mixer could 
be used for the third channel 
and the old scoring monitor 
room and mixer were em- 
ployed for the fourth channel, 
it would thus be possible to 
record with from one to four 
separate channels of centrally 
located apparatus. 

The rerecording of Class A de luxe musical pictures may be done 
with better perspective and balance in a room more nearly approaching 
the volume and acoustics of a large theater. It was also thought 
that rerecording activities could be expanded by installing a dubbing 
mixer on the new scoring stage. Thus by adequately equipping our 
facilities, all production schedules could be met. 

After the requirements for scoring and dubbing had been carefully 
analyzed, schematic diagrams were planned to provide complete flexi- 
bility without unnecessary equipment. Fig. 9 shows a block sche- 
matic of the scoring channel in a typical split-channel setup. 

A cabinet rack is provided adjacent to the mixer console; it con- 
tains 12 microphone preamplifiers and four single-stage booster ampli- 
fiers to be used for isolation amplifiers or where additional gain is re- 
quired in low-level circuits. One headphone amplifier (RCA MI- 
9354) is available which can be fed either from the vocal channel 
bridge buss for prescoring or from a soundhead for. postscoring. 
Metering facilities, in the form of a meter and rotary switch, are 
provided for all plate and filament circuits in the rack. 
The patch bays are divorced from the mixer console, and high- 



FIG. 9. Schematic of scoring channels (split). 

and low-level patch bays are located in this cabinet rack. The 
mixer console, placed as shown in Fig. 10, commands a full view of the 
stage and vocal room. The console is equipped with a 12-position 
mixer that may be split into two channels of eight and four each with 

FIG. 10. Scoring-mixer console. 



a master volume control. The schematic in Fig. 1 1 shows the mixer 
circuits and the keying arrangement for a split mixer or for combining 
both sections into a 12-position mixer. Each mixer position is pro- 
vided with three-position low-frequency attenuation of zero db, 5 db, 
and 7 db at 100 cycles. Each channel requires a volume indicator 
installed as shown in the photo, Fig. 12. One is the standard RCA 

FIG. 11. Schematic of scoring-mixer circuits. 

neon volume indicator and the second indicator is a vacuum-tube 
meter with fast attack timing (60 milliseconds) and slow delay timing 
(0.5 sec), which was developed by the Republic Sound Department. 
Volume-limiter ceiling controls for the two channels are located at the 
right side of the mixer panel, and on the panel to the left of the mixer 
panel are mounted the p. a. signal and intercommunication controls. 
Two-way p. a. communication is provided to the music director, the 
stage proper, the vocal room, the recorder, the projection room, and 
the film machine room. 

The p.a. microphone and speaker are mounted in the vertical sec- 
tion of the console in front of the mixer. The panels on the top sur- 
face of the console are pivoted for accessibility. Doors are provided 
on the ends and front of the console to expose the internal wiring to the 
local terminal blocks. 

Speech circuits located in the scoring stage are terminated in the 
cabinet rack. The transmission lines to the main amplifier room are 


composed of No. 19 twisted pair lead-shielded cables in 2-in. conduits 
which are placed low enough in the ground to give maximum safety 
against any future grading or ditching for underground lines. The 
conduks are laid between stages 9 and 12 and are approximately 225 
ft long. The conduits carrying circuits intended for a particular rack 
are terminated behind that rack. 

Fig. 13 shows the rack layout of the new installation. One new 
rack was added to the left end and six new racks were added to the 
right end. In amplifier bay No. 1 are two MI -9 3 28 amplifiers with 
a total gain of 100 db, which are used to increase the output of the 
MI -487 5 pickup to 0-db level for feeding the monitor amplifiers of the 
acetate playbacks. Nine phototube preamplifiers for the new 

FIG. 12. Scoring console. 

soundheads are so mounted on shelves that the plugs on the amplifiers 
line up with receptacles on the shelf assembly and engage when set in 
place. Two volume compressors for the dubbing channel are also 
included in this rack. A patch bay is provided where all speech cir- 
cuits in this rack and trunk lines to other points appear. 

Bays 2, 3, 4, and 5 were not changed except for the additon of trunk 
lines to other new bays. Bay 6 was allotted for additional circuit lab- 
oratory facilities and test trunks to all new amplifier bays, mixer con- 
soles, projection rooms, and other apparatus. On bay 7 is channel 
No. 3 or the normal dubbing channel for stage 12. This channel is 
also used for the second channel on split or dual recordings. In bay 8 
is channel 4 or the normal scoring channel. Bay 9 houses the monitor 
amplifiers and the decompensator for channel one. Bay 10 carries 



the monitor amplifiers, etc., for channel 2, and bays 11 and 12 contain 
the monitor amplifiers, etc., for channels 3 and 4, respectively. In 
bays 11 and 12 are also mounted two MI-9354 amplifiers, of which one 
is used for a headphone amplifier for the output from the orchestra 
channel bridge buss for split-channel operation; the other is used for 
orthoacoustic recordings. 

In bays 13 and 14 are the main signal system patch panel and the 
p.a. switching panels. The signal system may be patched to include 
only those positions (mixer, recorder, projection, machine room) which 

FIG. 13. Amplifier racks, new installation. 

will be required to operate together during a recording session and are 
not to interfere with any other independent unit operating on either of 
the two stages. The signal lights are "recorder ready", "machine 
room ready", "projection room ready", and the "running" lights. 
Thus, all positions operating as a unit have visual indication when 
all positions are ready and when on a "take". The running light cir- 
cuit also controls the red warning lights at each entrance of the stage. 
The p.a. switching panel selects p.a. circuits for the entire instal- 
lation consisting of 15 stations. Each station terminates at a desig- 
nated rotary switch which selects the desired p.a. amplifiers and buss, 



and also the relay keying circuits. Five p. a. amplifiers and busses 
are provided so that five different systems may work simultaneously 
with no cross talk. The selector switches of stations that are to com- 
municate together are set to the same -p. a. buss and nonassociated p.a. 
networks are set on other p.a. busses. This arrangement has proved 
very flexible and foolproof since the setup operation is rather simple. 
One standard film recorder and two Class B push-pull film recorders 
have been added to make a total of three Class B recorders and two 
standard recorders. Each recorder position has its associated p.a. 
signal and interlock control panel mounted in an inclined panel on 
the front of the recorder table. Behind each recorder, mounted flush 
in the wall, is the combination p.a. and monitor speaker with individ- 
ual volume control. 

FIG. 14. Schematic of playback system. 

An additional disk recorder provides facilities for two scoring ses- 
sions or for two acetate copies when needed. A transfer switch on the 
recorder control panel selects the record position or the payback posi- 
tion. Playback reproduction is through the normal monitor circuits 
and through an auxiliary mixer on the scoring stage Fig. 14 shows 
the block schematic of the playback system. 

All filament and plate requirements are supplied from rectifier units 
mounted in the power room. The filament supply units are dry-disk 
rectifiers and the plate-supply units are of the tube-regulated type. 
No line- voltage regulators are used since the main supply is fed from 



an isolated bank of transformers and the line voltage is constant 
within plus or minus 3 per cent. 

Two new Selsyn distributors have been added, making the total 
four, with space and load requirements allotted for a fifth, should it be 

The interlock selector switch 3 panel is located in the film machine 
room. A total of 30 Selsyn motor circuits located in both stages 9 
and 12 are controlled by the 30 selector switches which have five posi- 
tions so that one or more motor circuits may be connected to any of 
the four present distributors. The positions on the switch may 
be selected prior to closing the contacts, thus eliminating the danger 
of moving the switch past a live buss. The distributor control cir- 
cuits are patched to the selected control stations by patch cords. 

FIG. 15. Dubbing console and scoring stage. 

The dubbing console located on the new stage (Figs. 15 and 16) 
was designed to incorporate the features that were found desirable 
from past experience. A 12-position mixer was expanded into three 
panels of four positions each, physically so spaced to allow three mixers 
sufficient elbow room. The main dialogue equalizer panel located be- 
tween the center and right-hand panels may be operated either by the 
first or second mixer, or an extra man seated between the first and 



second mixer. The console provides room for four mixers, if that 
many are needed. A three-position auxiliary mixer is located be- 
twee*h the center panel and the left-hand panel and may be operated 
by the mixer at either panel. The auxiliary mixer is used for rever- 
beration or other effects. Soundhead amplifier outputs are bridged 
by zero-db-gain bridging amplifiers and fed to the auxiliary mixer pots, 
thence through a suitable amplifier to the speaker in the reverberation 
chamber. The reverberated signal is picked up by a microphone and 
fed to one of the main mixer pots, where it is mixed with the original 
for the desired effect. Fig. 17 is a block schematic. The remote con- 
trol for the door between the two reverberation chambers is located 

FIG. 16. Rear view of scoring stage. 

in a panel adjacent to the left-hand mixer panel and is calibrated in de- 
grees of door opening. Practically unlimited effects of reverberation 
may be obtained by the several variables, such as speaker placement, 
microphone placement, levels employed, varying door opening be- 
tween chambers, and equalization of both the signal fed to the 
speaker and the reverberated signal. Four RCA MI-10102 com- 
pensators are located within easy reach of the mixers and are used 
for auxiliary equalization. 

The main dialogue equalizer panel located adjacent to the right- 
hand mixer has six control knobs and will produce almost any type of 



compensated-frequency characteristic desired. A variable high-pass 
filter is provided with five different cutoff frequencies, namely, 80, 
100, 120, 135, and 150 cycles. A variable-dip equalizer is available 
with five resonant frequencies of 1500, 2000, 3000, 4000, and 5000 
cycles, having five step of 2 db each. The low-frequency attenuation 
is controllable in eight steps of 2 db at 100 cycles by rotating a knob 
counterclockwise from the zero position. Clockwise rotation from 
the zero position gives equalization in six steps of 2 db at 100 cycles. 
An attenuator with converse attenuation is connected to the same con- 
trol shaft and maintains constant insertion loss outside the equalized 
portion of the characteristics. High-frequency equalization is con- 
trolled by two knobs, one for the amount and the other for the fre- 
quency band. Equalization may be had at resonant frequencies of 

FIG. 17. Schematic of reverberation channel. 

3500, 5000, or 8000 cycles in six steps of 2 db each by clockwise rota- 
tion of the high-frequency control and by selecting the resonant fre- 
quency required. Constant insertion loss is also* maintained for this 
equalizer. High-frequency attenuation is provided by counterclock- 
wise rotation of the high-frequency control in six steps of 2 db each at 
6000 cycles. Curves are shown in Fig. 18. 

The total insertion loss of the complete equalizer is 24 db, which is 
compensated for by an RCA MI-11207A amplifier. Any one or all 
of the equalizers and filters may be patched through a key located on 
the panel and may be keyed in or out of the circuit for preset amounts 




: : : 4 * 

S I 5 5 


of compensation. Four ceiling controls for the compressors and lim- 
iters are located adjacent to the right-hand mixer panel. Duplicate 
p.a. and signal panels are located at each end of the console. All cir- 
cuits in the console appear in the patch bays located on the back of the 
console. Panel doors are provided in the ends and the back, giving 
easy access to all apparatus and terminal blocks. The panels are 
pivoted for ready inspection and maintenance. 

The review room, projection room, and music department offices 
are located on the second floor. The review room (Fig. 19) is 26 ft 
6 in. long, 15 ft wide, and 10 ft high, and carries acoustic treatment 
comparable to the main stage. Two-way corner speakers were used 

FIG. 19. Review room. 

to conserve space. A six-position mixer with VI, p.a. signal, and in- 
tercommunication shown in Fig. 20, is provided for music and effect- 
track checking or may be used as a third scoring channel or for re- 
recording of up to- six tracks. 

The projection room is 15 ft wide, and 26 ft 6 in. long, and 9 ft 
high, and treated with acoustic plaster. The room is equipped with 
fouf Simplex projectors complete with RCA MI-9066 soundheads 
and preview attachments, two additional RCA MI-9066 film repro- 
ducers, one RCA PG-142, and one RCA PG-140 reproducing system, 
rewind bench, and film storage. 



The PG-142 and two projectors were provided for the main stage 
to allow continuous showing, although the prime use of the stage is 
the scoring of music. Electric rewind attachments on the projectors 
eliminate lost motion during rewind operation. The soundhead out- 
put is normally connected for fader operation through the PG-142, 
but may be keyed through a preamplifier and sent to the scoring con- 
sole for cuing and through the headphone amplifier to the stage 

The preamplifiers and low-level patch bays are mounted in a cabi- 
net flush with the wall adjacent to the PG-142 amplifier rack. The 
p. a. signal and intercommunication panel is mounted flush in the wall 
between the two projectors. The PG-140, two projectors, and the 

FIG. 20. Mixing console of review room. 

two film reproducers are provided for the review room where continu- 
ous shows may be run. However, since the review room is used 
almost exclusively for checking music and effect tracks, the sound- 
head outputs are normaled to preamplifiers, whose output is fed to 
the review room mixer and thence to the PG-140 and review room 
speakers. Since the projectors are provided with preview attach- 
ments and two film reproducers are available, four tracks may be 
run in synchronism without additional facilities. 

The driving motors are Selsyn interlock, and an interlock distribu- 
tor panel is located in the booth. Thus, if every machine is available, 


it is possible to interlock them and to run the outputs from six tracks 
to the review room mixer. The Selsyn system may be reversed for 
stop, go, and reverse runnings. The reversing switch is incorporated 
in the lock switch, which is a three-position switch. One position is 
for lock on the foward-running position, the center position is off, 
and the third position is for lock in the reverse-running position. 
All take-ups are modified for reverse-running. Low- and high- 
level lines between the projection room and monitor room, dubbing 
console, and main amplifier room are provided. Because of the 
close contact required between the sound department and music 
department, both departments were installed in this building. 
The sound department has offices on the first floor, as has the 
music library. The music department offices are located on the 
second floor, and since the review room is also there, any picture 
reviewing necessary by the music department may be done without 
lost time. The third floor is reserved for future expansion. 

'Leading artists, such as Leopold Stokowski and Artur Rubinstein 
who have made recordings on this stage, praise its excellent acoustic 
qualities. Tremendous interest has been stimulated throughout the 
industry for similar structures, and other major studios have negoti- 
ated for the use of this scoring stage. 


1 LOOTENS, C. L., BLOOMBERG, D. J., AND RETTINGER, M.: "A Motion Picture 
Dubbing and Scoring Stage", J. Soc* Mot. Pict. Eng., XXXII, 4 (Apr. 1939), 
pp. 357-380. 

2 BLOOMBERG, D. J., AND LOOTENS, C. L.: "Class B Push-Pull Recording for 
Original Negatives", /. Soc. Mot. Pict. Eng., XXXIII, 6 (Dec. 1939), pp. 664-669.. 

8 BLOOMBERG, D. J., AND WATSON, W. O.: "A New Selsyn Interlock Selection 
System", /. Soc. Mot. Pict. Eng., 47, 6 (Dec. 1946), pp. 469-473. 



Summary. A brief summary of the development of the low-reflection coating 
is given. As instruments were improved, a point was soon reached where further 
improvement in optical design required elimination of surface reflections. An ex- 
planation of the color effect observed in a coating and a method of applying a film 
whose index varies from top to bottom which will eliminate the color effect is described. 
A comparison is drawn between the coating produced by natural aging, nitric acid, 
hydrofluoric acid, magnesium fluoride, and the American Optical Company low- 
reflection coatings. This shows that the magnesium fluoride is the most practical 
film to date when all factors are taken into consideration. The American Optical 
method of applying coatings without using vacuum is described. There are four main 
advantages in using a low-reflection coating: (1) elimination of reflections, (2) in- 
creased transmission, (3) increased contrast, and (4) increased chemical stability. 
Several tests are outlined which indicate that glass properly coated with magnesium 
fluoride will withstand more chemical abuse than uncoated glass. A new method of 
removing a high-temperature-baked low-reflection coating using melted crystals of 
potassium bisulfate is described which will materially speed the decoating process. 

Since the beginning of time, man has striven to improve the tools 
with which he has to work. The first telescopes were very simple and 
no thought was given to reflection losses because these difficulties 
were minor in comparison to the other shortcomings. To increase 
the speed and quality of various types of lens systems, more lens ele- 
ments were added, and each additional element decreased the total 
transmission of the system and decreased the contrast of the image. 
It was soon found that there was a practical limit to the number of 
air-spaced elements that could be used in a lens system without en- 
countering serious difficulties with reflections. This was true with 
camera lenses and to an even greater extent with the many com- 
plicated instruments required by the Armed Services. With each im- 
provement in optical design it became more obvious that the re- 
flection problem must be met and treated. 

* Presented Oct. 22, 1946, at the SMPE Convention in Hollywood. 
** Simpson Optical Manufacturing Company, Chicago, 111. 




Vol 49, No. 1 

Mr. H. Dennis Taylor, in 1892, was the first to recognize the need 
for reflection-reducing films. He and many. other investigators tried 
to treat lenses chemically, with only partial success. It was not until 
Dr. John Strong first used the vacuum process to apply coatings to 
glass that men began to see a practical way of reducing surface re- 
flections. He placed a quantity of calcium fluoride in a small heater 
and arranged the glass parts above. The assembly was enclosed 
within a bell jar and the air removed. The calcium fluoride was then 
vaporized by the heater, and a thin coating condensed on the glass. 
Other investigators made refinements on the basic idea, and by 1941 
the present process was well under way. 


FIG. 1 . The effect of light of 3 wavelengths on a layer of magnesium fluoride. 

The war emergency took hold at this point and in the short space of 
five years the coating of glass has become an industry. Today, prac- 
tically every optical company has coating facilities, and uses for the. 
process are being extended. 

Now let us look more closely at the low-reflection film. The most 
noticeable thing about a low-reflection coating is its color. The 
average individual usually remarks that the coating is purple. It is 
discouraging to the coating man that the only comment made on his 
accomplishment is to mention one of its disadvantages. Seeing that 
the color of the coating is so noticeable to the user, I would like to 

July 1947 



point out the reasons for this effect. Fig. 1 shows a cross-section view 
of a coated piece of glass. This glass has an index of refraction of 
1.69; and the magnesium fluoride an index of 1.38. The thickness of 
the coating is one quarter of the wavelength of green light. This 
amounts to a thickness of about 3.8 million ths of an inch. The wave 
character of light is represented by the wavy lines. In an uncoated 
piece of glass we would expect about 6.5 per cent reflection per surface, 
and the Balance transmitted as indicated by example A . In example 
C the reflected light is broken into two beams of about equal intensity. 
In this case the thickness of the magnesium fluoride film is one fourth 
of 5500 A thick. The beam of light reflected from the glass surface 
has been made to travel one-half wavelength farther before being re- 

o ru u> en o> - 




GLASS B&L- EDF-2- INDEX 1.689 



i 1 1 i 1 


1 1 1 | 

)0 500 600 70 


FIG. 2. The reflection values for a single surface of glass coated with 
magnesium fluoride. 

united with the first beam. The two beams will then cancel each other 
and the result is a reflection of 0.3 per cent. This canceled light must 
go somewhere, so it appears in the transmitted beam of light. There 
is another condition that is important in minimizing surface reflections. 
The two small beams of reflected light must be equal in intensity in 
order to bring the reflection to zero. In example C our result was 0.3 
per cent because the reflection from the upper surface was too large. 
To give zero reflection, the index of refraction of the film must be 1.3 
In case B we see what will happen to the same coating when violet 
light, 4000 A in length, is used. The second reflected wave, being 
shorter than case C, will no longer be exactly one-half wavelength out 
of phase with the first reflected beam. This will result in incomplete 
cancellation. Example D shows a similar condition when red light 



Vol 49, No. 1 

of about 7000 A is used. The second reflected beam, being longer 
than 'the one in example C, will result in incomplete cancellation and a 
reflection of about one per cent. 

From these three examples, it is seen that the color of the coating is 
caused by the variation of the wavelength of light found in the visible 
spectrum. Fig. 2 shows a typical reflection curve that could be ob- 
tained from a glass with an index of refraction of 1 .69 and a magnesium- 
fluoride film. Along the abscissa the wavelength varies from 4000 to 
7000 A. The ordinate is calibrated in percentage reflectance. You 
will notice that at 4000 A there is a reflection of 1.8 per cent, and as 
the wavelength increases the reflection drops to a minimum at about 
5500 A. The reflection increases again in the red. 


^ -^ - ^ i i 

i . . .. 



B &L - BSC-2 - INDEX 1 


= = 



~~~- _ 


, ,' , , 

1 1 1 I 


, , , , 


500 600 



FIG. 3. The reflection values for magnesium fluoride on BSC-2 glass. 

There are two main factors that must be observed to reduce reflec- 
tion to zero. A coating must be produced so that the two reflected 
beams of light are exactly out of phase with each other. In addition, 
the two beams must be of. the same intensity. If these conditions are 
not met, incomplete cancellation of light will result. The surface re- 
flection from an isotropic medium, such as our glass, varies with the 
index of refraction of the material. To make the two beams equal, it is 
necessary that the index of the refraction of the coating material be 
equal to the square root of the index of the glass. Thus, for a glass 
that has an index of 1.69, the coating material should be 1.3, Unfor- 
tunately, there is no suitable coating material with an index of 1.3. 
If a glass with an index of 1.52 were used, such as common window 


glass, the film necessary to produce zero reflection should have an in- 
dex as low as 1.23. Fig. 3 shows the reflectance value for a mag- 
nesium-fluoride film supported on borosilicate crown glass. This 
shows clearly the result of deviating from the square-root condition. 

Many attempts have been made to find a satisfactory substitute 
for magnesium fluoride. Calcium fluoride, strontium fluoride, lith- 
ium floride, sodium fluoride, cryolite, and many other materials have 
been worked on, but failed because of poor mechanical or chemical 
properties. Some manufacturers add a small amount of calcium 
fluoride to the magnesium fluoride, which does reduce its index slightly 
but again mechanical and chemical properties are sacrificed. 
Magnesium fluoride can be applied to glass in a spongy or porous form. 
By applying the coating at higher pressures than recommended, 
spongy films with an index as low as 1.2 can be obtained. These wipe 
off readily. 

Magnesium fluoride has, so far, proved to be the best coating mate- 
rial, when all factors are taken into consideration. To obtain the best 
results, however, care must be taken in the preparation of the glass 
and in the methods of application. The glass must be cleaned to re- 
move fingerprints, grease, and even dust particles. Every spot of dust 
remaining on the lens during the coating process will result in a small 
uncoated area. 

The equipment for coating is relatively simple. A mechanical vac- 
uum pump connected in series with the diffusion pump is required to 
secure the vacuum. This is connected to the bottom side of a large 
steel-base plate which has the appropriate electrical connections. 

The magnesium fluoride is placed on a small heater and the lenses 
are arranged about 15 to 20 in. above. A large lens heater is placed 
over the lenses and a bell jar is placed over the entire assembly, so 
that it is sealed tight against the metal-base plate. The lenses are 
heated to approximately 450 F while the pumps are securing the desired 
vacuum. After the vacuum has reached about 10-5 mm of mercury, 
we are ready to apply the coating. The magnesium fluoride is heated 
until it vaporizes, and being in a vacuum it streams out in all directions 
much the same as light radiates from a lamp. The fluoride condenses 
on the relatively cool glass in a thin, uniform layer. Any obstruction 
that might come between the lens and the fluoride heater would cast 
a shadow on the glass being coated. The thickness of the coating is 
controlled by observing the color of sample glass arranged adjacent to 
one of the lenses to be coated. 

32 STRICKLAND Vol 49, No. 1 

This, by no means, is the only way to apply low-reflection film to 
glass. The natural elements produced the first low-reflection films. 
These were formed by the action of moisture on the less stable glasses. 
They were usually spotty and very hard, and not so efficient as our 
present magnesium fluoride. Many attempts to reproduce these 
natural films were made by Taylor, Kollmorgan, and Wright. The 
most successful films of this type were made by Frank Jones, working 
at Bausch and Lomb. His films were uniform and hard, but not so 
efficient as magnesium fluoride. 

Another chemical method of producing low-reflection film on glass 
was developed by F. H. Nicoll of RCA. This type depends upon the 
action of hydrofluoric-acid vapors on glass. This process has advan- 
tages over other chemical methods, in that ordinary window glass can 
be treated in a reasonable length of time. The process can be applied 
to large plates of glass as well as small. The one disadvantage noticed 
with this film is that the coating is somewhat uneven. This is not 
serious, however, and possibly could be overcome. 

One of the most remarkable, and indeed the most likely process to 
challenge the magnesium fluoride, has been developed by Dr. Molten 
of the American Optical Company. A lens to be coated is cleaned 
and rotated about a vertical axis at a moderate speed. Several drops 
of solution are dropped on the lens and allowed to whirl off and dry. 
This produces a uniform coating on the glass and takes less than one 
minute. The process can also be extended to large plates of glass by 
dipping them in the solution. These coatings can also be applied by 
spraying or swabbing. Films of this type are hard enough to apply to 
eyeglass lenses and come very close to the efficiency of magnesium 
fluoride. This type of film looks much the same as magnesium- 
fluoride films and can be washed and cleaned without damage. It 
is not oil-sensitive and will stand the 24-hr humidity test and salt at- 
mosphere test of the Frankfort Arsenal Specifications 5 1-70-4 A . The 
films are attacked by strong alkalis but can resist acids admirably. 

In addition to this film, Dr. Molten has developed another type of 
coating which has the unusual property of having the same reflection 
effect on all glasses regardless of index of refraction. A single surface 
will reflect, after coating, about 3 /io of one per cent even on glasses 
with an index as low as 1.52. This film owes its unusual properties 
to the fact that the index of the coating material is varied from about 
the index of air at the air-coating surface to about 1.5 at the glass- 
coating surface. The spongy nature of this coating makes this film 


soft and oil-sensitive, but owing to its remarkable properties, man 
uses will be found for it. 

Efforts to replace magnesium-fluoride film with chemical films hav 
failed because nothing has been found with the abrasion-resistin 
quality of the fluoride. The usual test given coating by the Arme 
Services is to rub the coating with standardized erasers, */4 m - m dian 
eter, for 20 strokes across the same area of the lens at a pressure of 2.511 
This amounts to a pressure of approximately 50 Ib per sq in. If ther 
are no visible effects from this treatment, the element is considere 
acceptable. In addition to this test, Army and Navy requirement 
call for 24-hr immersion in salt water; a 95 F salt atmosphere for 2 
hr, and a humidity test at 120 F at 95 per cent relative humidity for 2 
hr. The magnesium fluoride can stand these and more severe test! 

I have had many requests for instructions on how to clean coate 
lenses. It would seem from the Army-Navy specifications, that an 
treatment short of a wire brush would be all right. I would, howevei 
treat a coated lens as I would any other high-quality optical elemenl 
I would recommend water or alcohol as a cleaning agent. A clea: 
handkerchief is excellent for drying the lens. If the lens is covere< 
with grit and dirt, try to remove this before real pressure with a hand 
kerchief is applied. 

Up until 1943, every effort was exercised toward making coating 
more durable in all respects. This development program was carrie* 
out so well that in October 1943, at an Army convention on coating 
it was decided that there was a need for a method to remove defectiv 
coatings. Several companies had accumulated moderately larg< 
stocks of lenses which had defective coatings, and if a satisfactor 
means could be devised, these lenses could be salvaged. Many peopl 
worked on this problem and finally the boric-acid process was de 
veloped. This consisted in boiling the lenses in a concentrated solu 
tion of boric acid and salt water for 2 hr. The process was not re 
liable and was very hard on certain types of glass. Dense bariun 
crown glass and some flint glass were strongly attacked by the solu 
tion. The Navy Department then devised a much better method 
which could be used on all types of glass. This consisted in heatin 
the lenses in a concentrated solution of sulf uric acid and boric acid f o 
about one hour. This process was very successful but required cau 
tion and considerable time was lost in decoating. 

Another process was developed recently at the Simpson Optica 
Company. This process consists in melting crystals of KHSO^ 

34 STRICKLAND Vol 49, No. 1 

potassium bisulfate, in a porcelain or enamel dish. The temperature 
should reach 250 C and care should be taken to prevent overheating 
and accidental addition of water. The lenses to be decoated are 
warmed and placed in the solution for approximately 2 min. The 
lenses are then removed and allowed to cool. The excess potassium 
bisulfate is then washed off in water. This process represents a con- 
siderable speed-up over the sulfuric-acid method and was found safe 
on all types of optical glass. 

There are four main advantages in using a low-reflection coating: 

(1) Elimination of reflection (3) Greater contrast 

(/) Increased transmission (4) Better chemical stability 

Elimination of surface reflections is mainly of interest to the photog- 
rapher. Considerable difficulty is experienced in photographing 
directly toward a spot of light. A set of ghost images is usually 
created under these conditions . Coating the camera lens will minimize 
these ghost images to a point where they are hard to detect. 

To demonstrate the increase in transmission obtained by coating, 
we coated all surfaces of one of our standard 5-in. anastigmat pro- 
jection lenses and compared actual light measurements with an un- 
coated lens of the same design. The increase in transmission over the 
uncoated-lens system represents 33 per cent. This agrees very well 
with the theorectical reflection values. 

Fig. 4 shows a graph giving the percentage reflection in relation to 
the index of the glass. From this we can determine the amount of 
light lost, provided we know approximately the index of refraction of 
this glass used in the projection lens. The first lens in this particular 
anastigmat has an index of about 1.62. According to the graph, we 
would expect 5.6 per cent reflection from the first surface. If we as- 
sume that 100 per cent of the light strikes the first surface, the trans- 
mission will be 94.4 per cent. The second surface will reduce this 
94.4 per cent figure by 5.6 per cent which gives 89.1 per cent. The 
second lens in this system has an index of 1.65, which represents a re- 
flection of 6 per cent per surface. If we continue the process of re- 
ducing the total transmission received by each surface, by the amount 
of loss, we arrive at a total transmission of 70.2 per cent. The low- 
reflection coatings are not 100 per cent perfect. On a glass that has 
an index of 1.62, we can reduce the reflection from 5.6 per cent to 0.9 
per cent. This represents a gain of 4.7 per cent. On the other type 
of glass used in this projection lens, the gain is even greater, being 5.2 

July 1947 



per cent per surface. Going through the projection lens, using the 
reflection values for coated glass, we obtained a transmission of 95 per 
cent. This means that the coated lens will transmit approximately 
35 per cent more light than the uncoated lenses. 

The increase in contrast of a coated lens has been investigated 
thoroughly by Tyler, Morris, and Jewett. They performed a series 
of photographic tests which showed that in photographing a subject 

1 . 1 *+ 






/ / 




: ROM 












/ / 

O 1.64 


/ / 





1 60 




TH / 

)RIDE / 




\ , 





\ * 



"J 1.56 




















1 50 













/ } 







0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8. 


FIG. 4. White-light reflection values from one surface of glass and th 

percentage gained by coating. 

which contains a large illuminated area such as the sky and darker 
foreground objects, a certain amount of the light from the sky is dou- 
bly reflected in the lens system and this light fogs over the darker areas 
of the picture. As expected, this contrast effect is much more notice- 
able when the light area of the picture is very bright. \ 

A series of tests was made at Simpson Optical Company to deter- 
mine whether a low-reflection film would protect glass from staining 
because of moisture. Two glasses were chosen which were known to 

36 STRICKLAND Vol 49, 

stain at a rapid rate. These were extra dense flint and barium c 
glasses. One side of each sample was coated with magnesium flu 
in the usual manner. The samples were placed in a salt-spray 
net in which we had replaced the salt solution with distilled v 
The samples were maintained at a temperature of 120 F for 1( 
The relative humidity inside the cabinet was 100 per cent and 
water accumulated on both sides of the samples during the test 
examination the plates were found to be stained only on the unc< 
side. The coated side was subjected to the regular Army-Na^ 
spection procedure and found to pass inspection satisfactorily, 
this we concluded that the application of the low-reflection co 
to these less stable glasses will retard the staining effect at le; 
long as the coating will stand up. Tests were also performed c 
more stable glasses such as borosilicate crown glass, but no st; 
effect was found on either the coated or uncoated areas. In addit 
this test, several coated pieces of glass were subjected to hydrofl 
acid vapors. An exposure that would completely frost an unc 
glass showed only small pit marks in the coated samples. 

While the low-reflection coatings have been known for alm< 
years, all the developments which have made it practical have 
within the last five years. Coating is really in its infancy, and ^ 
expect many new developments. In the next ten years, coatir 
probably advance to a point where the index of refraction of the 
does not affect the reflection results. The first attempt at th 
made by Dr. Molten, but his coating which accomplishes this is si 
soft to be practical. 

Although the color effect of coating is primarily caused by a 
tion of wavelengths of light, this difficulty too may be mini] 
The lens designer is now able to use more air glass surfaces in or 
secure better image quality without the fear of poor contrast ar 


JONES, F. L.: "Some Properties of Polished Glass Surfaces", /. Soc. M< 
Eng., XXXVH, 3 (Sept. 1941), p. 256. 

TYLER, J. E., MORSE, R. S., AND JEWETT, F. B., JR.: "On the Perform 
Optical Instrument with Coated Lens System", /. Opt. Soc. Amer., 32 
1942), p. 211. 

LYON, D. A.: "Practical Application of Metallic and Non-Metallic F 
Optical Elements", /. Opt. Soc. Amer., 35, 2 (Feb. 1945), p. 157. 

NICOLL, F. H.: "A New Chemical Method of Reducing the Reflet 
Glass", RCA Rev., VI, 3 (Jan. 1942), p. 287. 



mtnary. During the evolution of cinematography there has been a constant 
ndfor increased light from a single*source. Early attempts to meet this demand 
tnade by improvements in lamp projection optics and by increasing both size and 
* input of the light source. l In recent years a great deal of work has been done 
d increasing the intrinsic brilliancy of the high-intensity carbon-arc source, 
has resulted in a series of carbons known as super-high-intensity carbons. 2 ' 3 
paper will describe the requirements of the motion picture industry which 
'Jit about the production of these super-high-intensity carbons and will cover 
'.> of the development and design of carbon-arc lamps to burn them. The use of 
-high-intensity carbon-arc units in motion picture studios may properly be 
?i into (a) Process background projection; 4 (b) Set lighting. 1 

rocess Background Projection. Process background projection 
means whereby a stereopticon slide or a motion picture may be 
t-cted onto a translucent screen to form the background for a 
,vhich has been constructed on the opposite side of the screen, 
tographing both the projected image and the set results in a com- 
te picture. 6 Inasmuch as the set on the camera side of the screen 
laminated for proper photographic exposure, it is quite evident 
the screen light must be of a much higher level than that in a 
ion picture theater. 

r hen this process first came into use, standard motion picture pro- 
on lamps were the only available equipment. In attempts to 
?ase screen light the studios tried carbons from every available 
ee and were requesting carbons of higher current-carrying ca- 
ty than the design of the lamps permitted. At that time, car- 
> of higher current capacity were also of larger diameter and not 
] creased intrinsic brilliancy. Because the optical systems in use 
i filled by the smaller diameter, lower current-capacity carbons, 
l.ght gain was negligible and the increased heat often resulted in 
itisfactory operation. 

Presented Oct. 25, 1946, at the SMPE Convention in'Hollywood. 
Mole-Richardson Co., Hollywood, Calif. 




Vol 49, No. 1 

Higher levels of the proper intensity and quality of screen light 
called for co-ordinated effort between the studios and the various 
suppliers of equipment and materials. This demand resulted in 
activity on the part of the Research Council of the Academy of Mo- 
tion Picture Arts and Sciences which led to a co-operative investi- 
gation of the entire subject by all of the studios and manufacturers 
involved. 6 Subsequent to the investigation by the Research Council 
Process Projection Committee, a report was issued covering recom- 
mendations on process projection equipment. 6 

FIG. 1. Mole-Richardson test lamp. Close-up view through housing door 
opening showing details of the carbon-burning mechanism. 

The Mole-Richardson Company agreed to design and build a proc- 
ess projection lamphouse which would meet the requirements out- 
lined in the report. Inasmuch as a carbon-arc lamp is designed to 
feed and control carbons, this work was carried on in close co-opera- 
tion with the National Carbon Company, Inc. 

" A laboratory test lamp was designed and several of them were 
built for use in Mole-Richardson Company's and the National Carbon 
Company's research and development laboratories. This unit (Fig. 
.1) will accommodate any size positive carbon from 11 to 18 mm in 

July 1947 



diameter and may be adapted to burn other sizes. Separate motors 
control positive feed, negative feed, and positive rotation, so any 
desired variable of those three factors may be quickly obtained. The 
positive head may be adjusted for various lengths of carbon pro- 
trusion, and different types of air- and water-cooled positive carbon- 
current input contacts may be used. The negative carbon head is 
mounted on arms which allow it to be moved to carbon trim burning 
angles from coaxial alignment to 90 deg. Adjustments are provided 
for alignment of the carbons in both the horizontal and vertical 
directions. The negative head will accommodate various sizes of 


as t nfffes 



etoius of ctArett 

FIG. 2. Curves showing intrinsic brilliancy across the 
crater face of the "National" 16-mm super-high-inten- 
sity studio positive at 225 amp and the 16-mm high-in- 
tensity studio positive at 147 amp. 

air-cooled negative carbons, and can also be equipped with water- 
cooled negative carbon-current input contacts. This lamp makes it 
possible to obtain any set of operating conditions which may be 
desired for experimental work under conditions of very close control. 
As a result of a co-operative testing program, the National Carbon 
Company supplied a carbon trim consisting of a 16-mm X 22-in. 
super-high-intensity positive and a 17 / 32 - X 9-in. heavy-duty copper- 
coated negative to burn at a maximum current of 225 amp and 75 
arc v (Fig. 2) . This trim was chosen over others tested because of 
high intrinsic brilliancy, uniform distribution across the crater face, 
and steadiness of burning. For comparison, Fig. 2 also shows the 
standard 16-mm X 20-in. set-lighting carbon which is used in the 
M-R Type 170 lamp, and which delivers about one half the horizon- 
tal candle power. 



Vol 49, 1 

FIG. 3. M.-R. Type 250 process 
background projection lamp. Oblique 
view showing front and operator's 
sides. (The plate shown assembled to 
the front of lamphouse and the rheo- 
stat knob in lower left corner of control 
panel are parts of an associated process 
background projection equipment and 
are not furnished with the lamp.) 

The Mole-Richardson Type 
250 process projection lamp 
(Figs. 3 and 4), which burns this 
carbon trim, and its associated 
Type 251 grid (Fig. 5) have 
been designed and produced. 

The major features of the lamp 
design are briefly described as 
follows : 

(1) General construction. The 
outline dimensions of the lamp- 
house are such that it can be 

conveniently assembled with 
associated process projec 
apparatus. The front por 
of the lamphouse is arrangei 
accommodate the light-collec 
optical system together wit! 
supports and adjustments. L; 
hinged access doors are prov 
in the lamphouse and cor 
box for ease of maintena 

FIG. 4. M.-R. Type 250 pr 
background projection lamp. Yi( 
operator's side with lamphouse 
and control -panel door open, 
bank of three rheostats shown ii 
lower left corner of the control b 
part of an associated process 1 
ground projection equipment ai 
not furnished with the lamp.) 



FIG. 5. M.-R. Type 251 grid for 
Type 25O proccM background projec- 
tion lamp. Oblique view showing side 
and contactor panel end. 

um construction is used 
er possible so that the 
is kept to a minimum. 
using and its doors are of 
construction with 
wall fabricated of as- 
terial. This type of 
results in a low 
of heat and sound 
arc to the surround- 
. Extra space is pro- 
in the control box for 
instruments, switches, 
pch are used with the asso- 
ed process projection apparatus. The control box and lamphouse 
be conveniently separated for ease of handling and shipment. 
I) Positive carbon control. The positive carbon is rotated con- 
so that an even crater is maintained. A photronic-cell con- 
c causes the positive carbon to be fed forward as it burns so 
the source of light is maintained within very dose limits at the 

: the light-collecting optical system. 

u/tt* carbon control. The desired arc length is continuously 
closely maintained by a control circuit which positions the nega- 
carbon. When the arc switch is turned on, the negative carbon 
used to be fed forward until it contacts the positive carbon, thus 
;ig the arc. The control mechanism then immediately 
he negative carbon to a position corresponding to the proper 
gth, and maintains this arc length by continually feeding the 
forward as it burns. 

) Cooling. The lamp is designed for satisfactory operation 
t forced ventilation so that the objectionable noise of a ven- 
fan is absent. Openings in the lamphouse are provided for 
draft, and are arranged in such a manner that the resulting 

do not interfere with the stability of the arc. 
water-cooled positive head encloses carbon -con tact brushes 
fen are cooled by their contact with a water-cooled casting. The 
Ming water is circulated through the casting between the arc and 
I brushes, so that the operating temperature of the brushes is con- 
erably lower than in conventional designs, and it is expected that 
tie or no brush maintenance will be required. 

42 HANKINS Vol 49, No. l 

A water-flow indicator is located on the rear of the lamphouse. 

(5) "Douser". A "douser" in the form of a metallic plate is pro- 
vided, which can be swung into position between the positive carbon 
and the light-collecting optical system. Its motion is mechanically 
interlocked with the motion of the operator's lamphouse access door. 
Closing or opening the door causes the douser to assume its position 
between the positive carbon and the optical system. Hence after the 
door is closed, the douser will protect the optical lens from heat shock 
and hot particles caused by striking of the arc. When the door is 
opened, the douser protects the lens from thermal shock which might 
result from cool air entering the housing from the outside. Manual 
positioning handles are- located external to the lamphouse, so that 
the operator can "turn on". or "douse" the light through the optical 
system while the arc continues to burn. 

(6) Control panel. The control panel is equipped with instruments 
for indicating the line voltage, arc voltage, arc current, and length of 
unburned positive carbon. An arc-image screen provides the 
operator with a calibrated visual indication of the positions of the 
carbons. Knobs are provided for setting the arc-length regulating 
circuit, and for manual adustment of the positive and negative carbon 
positions. The lamp operation is entirely automatic, and is con- 
trolled by a small "off -on" toggle switch located on the control panel. 

The M-R Type 251 grid which is supplied with the process pro- 
jection lamp is designed specifically for the application. Adequately 
ventilated grid resistor units, which carry the arc current and produce 
the required voltage drop, are positioned in the center of the unit. 
Selector switches are mounted on a switch panel on one end of the 
unit, with connections made to various taps on the grid resistors. 
By manipulation of these switches, satisfactory arc operation can be 
attained with arc currents of 150, 180, 200, or 225 amp with any line 
voltage of 110 to 130 v in 5-v steps. 

Two line contactors, a starting contactor, a time-delay relay, an 
auxiliary relay, and a selenium rectifier are mounted on the contactor 
panel on the end of the unit opposite to the switches. The coils of 
the main-line contactors are connected across the supply through 
the arc switch on the lamp-control panel, in series with the selenium 
rectifier. The rectifier prevents the main-line contactors from being 
energized if- the supply to the system is not of the correct polarity. 

A "starting resistance" is provided in the grid circuit to limit the 

July 1947 CARBON-ARC LAMPS 43 

current on arc strike, and hence prevent the positive carbon crater 
from being damaged by the initial thermal shock. The starting re- 
sistance is automatically cut out of the circuit when negative carbon 
has retracted to its approximate operating position. This operation 
is accomplished by time-delay auxiliary relay and starting contactor. 

Two bus bars are provided on the grid for connection to the direct- 
current supply. The grid is equipped with three cables for connec- 
tions to the lamp, two heavy single-conductor cables for conducting 
the arc current, and one small three-conductor cable for the control 
circuits. The unit is mounted on sturdy rubber-tired casters of 
large diameter for portability. 

The above-described process projection lamp and grid combination 
is representative of the present-day knowledge in the art of producing 
this particular type of projection equipment. However, research and 
developmental work is continually being conducted with efforts 
directed toward more and steadier light. Experiments are being 
made in connection with the possible use of a small-diameter water- 
cooled graphite negative carbon, which may produce a steadier light 
than is produced with the larger air-cooled negative carbons, and with 
no loss of light. 

Tests are being made on a brushless water-cooled positive carbon 
contact unit without moving parts, which also promises to contribute 
toward the steadiness and increase in light output. Positive carbons 
are being considered which have brilliancies of as high as 1400 candle 
power per mm 2 and which may be operated up to 400 amp. 

Set Lighting. A careful study of cinematographic technique 
indicates that the present-day cinematographer often strives for 
an illusion of a "one-source" lighting, particularly in medium and 
long shots. While he must use a large number of units for balance, 
modelling, back light, and other effects which indicate his indi- 
viduality, he works for an over-all result suggesting that the illumi- 
nation is coming from one source of tremendous brilliancy such as is 
found in nature when the sun is just at the right position. This 
effect may only be obtained with a high-brilliancy source sufficiently 
small in area to cast well-defined shadows. The shadows cast by 
the other units are either covered by the main source or are elimi- 
nated with fill light, and while there may be a hundred lamps on the 
set, all noticeable shadows are cast by the main source, creating the 
illusion of a one-source lighting. 

Previous to the advent of the super-high-intensity studio-type 



Vol 49, No. 1 

FIG. 6. M.-R. Type 450 super-high- 
intensity arc spot lamp. Oblique front 
view showing 24-in. diameter Fresnel 
lens and operator's access door. 

carbon there were three general 
types of carbon arcs available 
for the motion picture studios: 7 
(1) the low-intensity carbon arc 
where the principal light source 
is incandescent solid carbon at 
or near its sublimation tempera- 
ture; (2) the flame arc where 
the light source is the entire arc 
stream made luminescent by the 
addition of flame materials; (3) 
the high-intensity carbon arc 
where, in addition to the light 
from the incandescent crater 
surface, there is a significant 
amount of light originating in the 
gaseous region immediately in 
front of the carbons as the result 
of the combination of high cur- 

rent density and an atmosphere 
rich in flame materials. 

The low-intensity carbon arc 
has no present use in motion 
picture studio set lighting. The 
flame carbon arc is used in gen- 
eral lighting units for front light, 
fill light, and to illuminate back- 
ings. The high-intensity carbon 
arc is used in spotlamps. 

The M-R Type 170, operating 
at 140 to 150 amp and 64 to 67 
arc v, has been the most popu- 
lar carbon -arc lamp for use in 
creating a one-source lighting 
effect and for boosting daylight 
on exteriors. 1 However, the 
rather high light levels used on 
color pictures indicated a need 

FIG. 7. M.-R. Type 450 super-high- 
intensity arc spot lamp. Oblique rear 
view showing operator's control panel 
on rear of control-mechanism housing. 

July 1947 



for a unit of still greater volume 
and penetrating power. 

A demand on the part of 
directors of cinematography for 
higher-powered sources resulted 
in some attempts by the studios 
to adapt the 16-mm X 22-in. 
super-high-intensity carbon to 
the M-R Type 170 lamp. The 
same troubles were encountered 
that had plagued the process pro- 
jection departments when they 
attempted to increase current in 
standard projection lamps be- 
yond the design characteristics. 

When a carbon trim is burned at 225 amp in a Type 170 lamp- 
house, the interior of the unit becomes overheated, endangering the 
carbon-feed motor-current leads, and positive carbon brushes. A 
carbon trim which will burn steadily under conditions of proper 
lamphouse ventilation may become erratic and unsteady if the 
control mechanism is overheated. The gear ratios controlling the 

FIG. 8. Arc element and control 
mechanism subassembly for M.-R. 
Type 450 super-high-intensity arc spot 
lamp. Front oblique view showing 
unit removed from lamphouse. 

FIG. 9. Chart indicating relative illumination char- 
acteristics of M.-R. Type 450 lamp burning the 16-mm 
super-high-intensity studio positive at 225 amp and M.-R. 
Type 170 lamp burning the 16-mm high-intensity studio 
positive at 150 amp. 



Vol 49, No. 1 

carbon-feed rates do not correspond to the burning rates or the 
burning-rate ratio of the higher-current carbons. 

To meet the demand for a higher-powered unit, the M-R Type 
450 lamp was designed (Figs. 6 and 7). This unit is equipped with a 
24-in. diameter Fresnel-type condenser lens. The drum is of suf- 
ficient diameter to ensure proper ventilation, the unit is wired for the 
increased current, and the feed motor, feed-motor rheostat, arc 
switch, and pin-plugs are located in a separate compartment on the 






007" ^ 



> LAt 
























. rxx 











10 10 90 40 SO GO 70 80 90 /OO //O 

FIG. 10. Typical curves showing illumination at center 
of 20-ft diameter spot at various distances from M.-R. 
Type 450 lamp burning the 16 -mm super high -intensity 
studio positive at 225 amp and M.-R. Type 170 lamp 
burning the 16-mm high-intensity studio positive at 150 
amp. (The diameter of the spot is defined as the diameter 
at which the illumination is 10 per cent of the maximum 
illumination present at the center of the spot.) 

back of the lamp in order to limit their operating temperature rise. 
This separate compartment and the lamphead mechanism can be 
removed from the lamphouse as a unit (Fig. 8). Hence, a subas- 
sembly of the working parts can be set up on a bench for convenient 
servicing. The carbon trim consists of a 16-mm X 22-in. super- 
high-intensity MP studio positive and a 17 /32- X 9-in. Hff cgred 
Orotip negative burning at 225 amp and 75 arc y. 

July 1947 CARBON-ARC LAMPS 47 

The chart in Fig. 9 gives a comparison of the illumination char- 
acteristics of the Type 450 and Type 170 lamps. In the maximum 
flood condition, the amount of luminous flux in the Type 450 beam is 
approximately double the amount in the Type 170 beam, and the 
apparent horizontal beam candle power is almost tripled. In the 
minimum spot position, both the flux and candle-power values for 
the Type 450 lamp are approximately twice those for the Type 170. 

A comparison of the illumination of a 20-ft diameter spot as pro- 
duced by the Type 450 and Type 170 lamps is shown in Fig. 10. It 
is apparent that the Type 450 lamp represents a considerable increase 
in the "penetrating power" of lamps for studio-set lighting. 

Another unit in the advanced stages of design is a super-high-in- 
tensity spot projector which will be similar to the Type 450, but which 
will be equipped with an integral optical system for throwing a well- 
defined and closely controlled spot for use in follow shots such as 
would be made in a skating picture. 

Experience gained in the manufacture of specialized searchlight 
equipment during the war will be of considerable value in increasing 
further the light output of motion picture studio lamps using super- 
high-intensity carbons. 

We wish to acknowledge the co-operation of the Transparency De- 
partment and Mr. Farciot Edouart of Paramount Studios in the 
design and production of the special process lamp; the splendid co- 
operation of numerous cinematographers and the Electrical Depart- 
ment members in the work which was done on the "Brute" Type 450 
lamp ; and the National Carbon Company, Inc. 


1 LINDERMAN, R. G., HUNDLEY, C. W., AND RODDERS, A.: "Illumination in 
Motion Picture Production", J. Soc. Mot. Pict. Eng., XL,-6 (June 1943), p. 333. 

2 JONES, M. T., LOZIER, W. W., AND JOY, D. B.: "New 13.6-Mm Carbons 
for Increased Screen Light", /. Soc. Mot. Pict. Eng., XXXVIII, 3 (March 1942), 
p. 229. 

3 JONES, M. T., ZAVESKY, R. J., AND LOZIER, W. W. : "A New Carbon for In- 
creased- Light in Studio and Theater Projection", 7. Soc. Mot. Pict. Eng., 45, 6 
(Dec. 1945), p. 449. 

4 JOY, D. B., LOZIER, W. W., AND NULL, M. R.: "Carbons for Transparency 
Process Projection in Motion Picture Studios", /. Soc. Mot. Pict. Eng., XXXIII, 4 
(Oct. 1939), p. 353. 

6 EDOUART, F.: "The Paramount Transparency Process Projection Equip- 
ment", /. Soc. Mot. Pict. Eng., XL, 6 (June 1943), p. 368. 

* Research Council of the Academy of Motion Picture Arts and Sciences : 
"Recommendations on Process Projection Equipment", /. Soc. Mot. Pict. Eng., 
XXXII, 6 (June 1939). p. 589. 

1 MACPHERSON, H. G. : "A Suggested Clarification of Carbon Arc Terminology 
as Applied to the Motion Picture Industry", /. $oc. Mot. Pict. Eng., XXXVII, $ 
(Nov. 1941), p. 43Q, 



Summary. A new light modulator, recently developed, has very low distorts 
and greatly improved performance characteristics. It is of the magnetic type and \ 
mechanically and optically interchangeable with the present RCA sound-recordi 
galvanometers. The power required for 100 per cent modulation is 1.25 w. D- 
tortion characteristics, frequency-response curves, and impedance data are show 
The effect of bias current upon the performance characteristics is also given. 

A galvanometer of the magnetic type was first used with the RC 
studio recording optical system in 1932. Although the basic design . 
this galvanometer has remained the same since that time, its perfom 
ance and reliability have been considerably improved as the resu 
of changes in the mechanical, magnetic, and electrical component 
For many years the most serious obstacle to further reduction in dii 
tortion and hysteresis was the limit on galvanometer sensitivity s\ 
by the maximum power available from recording amplifiers already : 
use. This obstacle was eliminated when the decision was made to dj 
velop a new 10-w, high-quality recording amplifier, and to develop! 
new galvanometer to operate with this amplifier. The characterii 
tics of the new amplifier are described in a separate paper written c 
Kurt Singer. It is the purpose of the present paper to describe tl 
changes which have been made in the recording galvanometer and J 
show how these changes have reduced distortion and greatly inprovc 1 
the performance characteristics. 

The types of distortion which are most detrimental to the operatic 
of the recording galvanometer are : 

(1) Odd harmonic distortion of the wave shape 

(2) Even harmonic distortion of the wave shape 

(3) Hysteresis 

(4) Lack of linearity between current and deflection 

* Presented Oct. 25, 1946, at the SMPE Convention in Hollywood. 
'* RCA Victor Division, Radio Corporation of America, Camden, N. J. 



[feet of d-c bias current upon a-c modulation characteristics 

istortion of the frequency characteristic. 

.armonic distortion results almost entirely from magnetic satu- 

the armature A shown in Fig. 1 . This can best be reduced 
ining an armature with a reserve of flux-carrying capacity 
;han that required to produce the normal 100 per cent deflec- 
the mirror. Even-harmonic distortion is caused almost en- 
a dissymetry in the two magnetic paths, shown by the dotted 
? ig. 1 . This may be caused by the armature being off center, 
y be caused by the pole pieces B of Fig. 1 not being mechani- 

magnetically identical. The distortions listed above as 
(5) are caused largely by the inherent properties of magnetic 



FIG. 1. Top view of galvanometer. 

5. After these types of distortion are minimized by the use of 
magnetic materials, properly annealed, a further reduction is 
by increasing the length of the gaps C and D (Fig. 1), thereby 
the iron path with more air. Distortion of the frequency 
ristic may result from having the wrong amount of damping 
echanical vibrating system, or it may be caused by a variation 
: optimum ratio of inductance to resistance in the electrical 

ntioned above, one way of reducing the undesirable effects of 
3 make the magnetic circuit include more air and less iron, 
tew galvanometer this was accomplished by increasing the 
gaps C (Fig. 1) from their former value of 5 mils each to a 
10 mils each. The thickness of each nonmagnetic spacer D 
ck of the pole pieces was also increased from the former value 
s to a value of 50 mils. Taking into account the larger area 
ack gaps, the above change resulted in a 3-to-l increase in 

50 DlMMlCK Vol 49, No. 1 

the air reluctance included in the magnetic circuit. Even with the same 
armature and pole pieces used in former designs, the above change 
would have improved the distortion of the types listed above as (3), 
(4), and (5). We needed, however, to reduce the odd-harmonic dis- 
tortion as well, and this required an armature that had a greater flux- 
carrying capacity. This was accomplished in the new galvanometer 
by increasing the width of the vibrating section of the armature by 50 
per cent. The dotted lines of Fig. 2 show the former width, while the 
full lines show the new armature shape. The wider section is main- 
tained out to the point at which the armature enters the front air 
gaps. From here it tapers down "to its former width at the base of the 
knife edges. The thickness of the new armature is the same as in the 
former design, and the armature material is an iron-nickel alloy. 



FIG. 2. New armature. 

This change in armature dimensions allows 50 per cent more flux to 
enter the gaps before saturation is reached. But the change also in- 
creases the stiffness of the armature by nearly 50 per cent. The in- 
creased force necessary to overcome this stiffness is obtained by making 
the pole pieces of an iron-cobalt alloy. The use of this alloy allows 
the permanent polarizing flux density to be raised almost 50 per cent 
above that which is obtainable with the present pole pieces. In a 
balanced armature arrangement like that of Fig. 1, it can be shown that 
the force exerted on the armature is directly proportional to the total 
flux entering the air gaps (from the armature) and to the permanent 
polarizing flux density in the air gaps. 

With the new pole pieces, the amount of armature flux required to 
produce 100 per cent deflection of the wider armature is equal to that 
which was formerly required to produce the same deflection of the 
narrow armature. The wider armature and new pole pieces therefore 
give us the reserve flux-carrying capacity necessary to reduce mate- 
rially the amount of odd-harmonic distortion. 

In a magnetic circuit containing both air and iron, the effective 


value of hysteresis is reduced as the ratio of air to iron is increased. 
In the new galvanometer the ratio of air to iron included in the mag- 
netic circuit has been increased 3 to 1, and the effective hystersis- 
is therefoie substantially less than in the former design. The most 
important effect of this is that when the new galvanometer is used on 
a recording optical system, the light beam returns to its zero position 
(or its biased position) with a greater degree of precision. 

As mentioned abo\ e, the increase in gap lengths resulted in almost 
a 3-to-l increase in the reluctance of the magnetic path. This 
had to be offset by an equivalent increase in ampere turns in the coil. 
If the coil design were left unchanged and the current increased 3 
to 1, the necessary flux could be obtained, but the heat developed in 
the coil would go up nine times, and this would be prohibitive. To 
get around this difficulty, the amount of copper in the modulation 
coil was increased about 5 to 1, thus greatly increasing the coil 

FIG. 3. Circuit of new galvanometer. 

efficiency. Increasing the amount of copper in the coil also increases 
the ratio of inductance to resistance, and it is, therefore, necessary to 
add an appropriate amount of series resistance in order to obtain the 
required frequency response. This series resistance may be located 
in the amplifier, thus allowing most of the heat to be dissipated out- 
side the galvanometer. 

All galvanometers of the present design have had two separate coils, 
a modulation coil and a bias coil. In the new model, it was decided 
to make a single coil serve for both modulation and bias currents. 
This does not interfere with the normal operation of the bias-type 
noise-reduction system because the modulation current is maximum 
when the bias current is minimum, and vice versa. 

The coil for the new model galvanometer was made by winding 
enamel wire on a bakelite-coil form. A tap was taken off for the por- 
tion of the coil used for modulation, while the whole coil is used for 
bias. The determining factor in deciding the number of turns on the 
modulation portion of the coil was the output impedance of the new 



Vol 49, No. 1 

recording amplifier. The modulation portion of the coil was made to 
have an impedance which would assure galvanometer operation di- 
rectly out of the amplifier without the customary matching transformer. 
The whole coil was wound with enough turns to provide a bias 
sensitivity of 30 ma. This makes it possible to use the new galvanom- 
eter with present noise-reduction amplifiers, modified for bias-type 
noise reduction. It also works equally well when using the shutter- 
type of noise-reduction system. 

Fig. 3 shows how the new galvanometer is connected into the cir- 
cuit of the recording amplifier and the noise-reduction amplifier. 
Resistor RI is the series resistance referred to above and at present is 
located in the amplifier. The capacitor C\ is for the purpose of block- 
ing the passage of bias current into the recording amplifier. It has a 
value of 10 pf which results in an attenuation in the frequency re- 
sponse of only 0.5 .db at 60 cycles. The capacitor C 2 and the resistor 
R 2 are located inside the galvanometer case, and the series network is 
connected across the bias winding. The purpose of this is to improve 
the frequency response by partially neutralizing the inductance of the 
modulation coil in the mid-frequency range around 3000 cps. 

The vibrating system of the new galvanometer is damped by means 

of a line of tungsten-loaded 
neoprene, one end of which is 
fastened to an extension of the 
mirror plate, and the other end 
is left free. As the mirror is 
rocked, the end of the line is 
deflected in torsion, and at 
high frequencies (5000 to 
10,000 cycles) a relatively 
large amount of energy is 
radiated down the line. The 
line is long enough and its 
attenuation high enough for 
this energy to be practically 
all dissipated before it gets 
back to the mirror plate after 
being reflected at the open end 
of the line. Fig. 4 shows the 
DAMPING LINE- new damping line in relation 
FIG. 4. Galvanometer with damping line. to the other parts of the 








pp 1 

_ _ __ 


xi me. rr.iu4C.MTnn To meis 


FIG. 5. Details of mirror plate and damping line. - [ 

galvanometer, and Fig. 5 is an enlarged view of the mirror plate show- 
ing the method of fastening the line. The line is made of neoprene 
which is loaded with fine tungsten powder. Neoprene is used in 
place of rubber because it has a higher coefficient of damping and is 
less affected by temperature. Tungsten powder is used because it is 
very heavy and is relatively inert. 

The new line-type damper has many advantages over the antireso- 
nant type of damper which has been in use on RCA recording galva- 
nometers for many years. The line damper is easier to make because 
it is not tuned and is not critical as to length or size. It is easier and less 
expensive to service because it is a part of the bridge assembly and can 
be removed or replaced without affecting the armature and pole-piece 
assembly. It is relatively unaffected by temperature within the 



"IF F 


31 1 








6 , 


i lzo 








O goo 




2 ^ 






*. * 






> 20,00 


FIG. 6. 



Vol 49, No. 1 








_ _ * *~ 


- = = 
















3 on 








1 00 



FIG. 7. 

normal operating range. The damping is linear with amplitude, thus 
making it possible to obtain practically the same frequency response 
at all input levels. 

Fig. 6 shows how the impedance of the modulation winding varies 
with frequency. This curve was taken at the terminals of the galva- 
nometer. The dotted curve in Fig. 7 shows how the total impedance 
(galvanometer plus series resistor) varies with frequency. The full 
line of Fig. 7 shows the current required for 100 per cent modulation 
at different frequencies. The product of the two curves in Fig. 7 




FIG. 8. 

July 1947 



-48 -46 -44 -42-40 -38 -36 34 -32 -30 -28 -26 -34 -22 -20 -18 -16 -14 

FIG. 9. 

gives the voltage required to produce 100 per cent modulation at dif- 
ferent frequencies. By taking the reciprocal of the above products and 
expressing them in decibels we obtain the frequency response curve of 
the new galvanometer shown in Fig. 8. The dotted line in this figure 
shows the effect of a 10-/-rf capacitor in series with the modulation 

Fig. 9 shows input-output curves taken at a frequency of 1000 cycles 


100% MOCK 
3RD HARfc 



Z 18 

h! iff 




K 5f 



r( 06 









- _ 







.. . 










FIG. 10. 


Vol 49, No. 1 

id over a range of 40 db below 100 per cent modulation. The curve 
arked "with bias" was taken under conditions similar to those which 
)uld exist if a noise-reduction amplifier were used in connection with 
e galvanometer. As the signal was decreased, enough bias current 
is applied to keep one edge of the recording light beam coincident 
th one of the limiting lines. 

Fig. 10 shows how second- and third-harmonic distortion varies with 
iquency. These curves are taken at 100 per cent modulation at all 
?quencies. Fig. 1 1 shows how second- and third-harmonic distortion 
ries with amplitude. These curves were taken at a frequency of 
00 cps. 





























-* > 

1 "" 






FIG. 11. 

In addition to having lower distortionand better operatingcharacter- 
:ics, the new galvanometer is relatively free from temperature drift 
id from variations in sensitivity resulting from temperature. Tem- 
:rature drift has been reduced by clamping the pole pieces on a stain- 
>s-steel support having nearly the same temperature coefficient of 
pansion as the pole pieces. Temperature drift has been reduced 
ill further by cutting the armatures from a solid bar of nickel-iron 
Loy instead of obtaining them from rolled-sheet stock as was for- 
erly done. Apparently the rolling process results in strains which are 
)t completely removed by annealing. 
The new galvanometer is equipped with a pair of alnico magnets 

July 1947 ZOOMAR LENS 57 

which have been fully charged and then demagnetized to about half 
their maximum strength. When used in this way, these magnets are 
extremely stable. This is important from a service standpoint, since 
the magnets may be removed from . the pole pieces and later replaced 
without changing the air-gap flux density from the optimum value set 
at the factory. These magnets are also relatively unaffected by tem- 
perature, vibration, and stray fields. 

The galvanometer is mounted in a cylindrical aluminum case and is 
mechanically and optically interchangeable with the present RCA re- 
cording galvanometers. The capacitor which was formerly mounted 
on the back of the case has now been mounted inside the case. The 
glass window in front of the mirror has been made larger in diameter 
so as to allow light to enter and leave the galvanometer at a greater 
angle to the normal. The power required for 100 per cent modulation 
of the new galvanometer is 1.25 w. 

Acknowledgment is due J. L. Pettus and H. E. Haynes for im- 
portant work in connection with the development of the new re- 
cording galvanometer. 



Summary. The "Zoomar" lens is a varifocal objective for motion picture 
cameras which achieves the change of focus by the linear movement of a single barrel. 
The new feature of this lens consists of the principle of changing the focal length of 
the system by one group of lens components without consideration of the displacement 
of the image plane, while a second lens component, rigidly coupled to the first by the 
common barrel, compensates for this displacement. 

The "Zoomar" varifocal objective has been developed as a tool for 
making "zoom" shots in places and on occasions where the usual 
methods of wheeling the camera toward or away from the object are 
either impossible or uneconomical. On the screen it seems as if the 
camera were moving toward the object in reality the camera, as 
well as the object, remains stationary. The apparent movement is 

* Presented Oct. 25, 1946, at the SMPE Convention in Hollywood. 
** Research and Development Laboratory, 381 Fourth Ave., New York. 

58 BACK Vol 49, No. l 

produced optically. During the zoom the image size varies but is 
always in focus. The light transmission remains constant over the 
entire zoom. 

The basic differences between the Zoomar lens and other varifocal 
objectives has been discussed in a previous paper. 1 

Fig. 1 shows the Zoomar lens mounted on a Cin Kodak Special. 

Fig. 2 shows a cross section of the lens. When the zoom lever is in 
the forward position the Zoomar is a wide-angle lens with an equiva- 
lent focal length of 17 mm. As the zoom lever is moved back the 
equivalent focal length of the lens increases to 53 mm, and in the rear 
position it has the characteristics of a telephoto lens. 

FIG. 1. Zoomar mounted on Cine Kodak Special. 

The construction of the lens is mechanically simple with no gears or 
cams and only one movable barrel. The barrel carries five lenses 
which move with it, varying the size of the image and compensating 
for distortion and aberrations. The four lenses at the front end of 
the movable barrel produce the variation in the image size. 

The rear lens of the movable barrel keeps the image in focus during 
the zoom. The rest of the lenses are stationary and fixed to the 
Zoomar body. 

The principle on which the function of these four lenses is based, 
was first used in a varifocal viewfinder developed for the combat 
cameras of the Armed Services. 2 

July 1947 



FIG. 2 A . Wide-angle position. 

FIG. 2B. Intermediate position. 

FIG. 2C. Telephoto position. 

Changing the front lens increases the range of Zoomar from 35 to 
106 mm equivalent focal length, but the image remains in sharp focus 
and the light transmission is constant. 

Naturally, the complicated optical system of the Zoomar was 
basically afflicted with many aberrations. Correction of these 




H mm 



FIG. 3. 



Vol 49, 1 

aberrations was one of the major tasks in designing the Zooi 
Ordinary correction methods of optical design broke down and 
ways had to be devised. 

Fig. 3 is a graphical presentation of the spherical aberration 
coma at maximum and minimum magnification. The solid 
represents the spherical aberration; the dotted line shows c 
expressed, as is customary, offense against the sine condition, 
sine condition is fulfilled when the spherical and coma curves coin< 




*imm o -imtn -nmm o -imm 




FIG. 4. 

At minimum magnification the zonal spherical aberration is gres E 
giving a certain over-all softness to the picture. The sine curv 1 
to the right of the spherical curve which signifies negative corns 

At maximum magnification the sine curve lies to the left, indie .i 
positive coma. Therefore, the absolute amount of this most o j 
tionable of all aberrations does not increase. At the same tim 1 
zonal spherical aberration has its minimum when the magnific ;i 
reaches its maximum. 

Fig. 4 shows astigmatism and curvature of field. At mini;i 
magnification, that is, at the wide-angle position, the field is fla i 
the astigmatic difference is negligible. At maximum magnific [i 

uly 1947 



re have a slightly curved field; and, though astigmatic correction is 
chieved at the intersection of the two curves, we still have astigmatic 
ifferences above and below the node. 

If we compare astigmatism and spherical aberration we see that at 
he wide-angle position the correction of zonal spherical aberration 
ad to be sacrificed in favor of astigmatic correction, giving a flat, if 
3rnewhat soft, image. At maximum magnification where the smaller 
r gular extent of the field renders astigmatism relatively harmless, the 
tress has been put on spherical correction. 



2'/ 3 ' 


FIG. 5. 

Pig. 5 shows the distortion of the system which is negligible. 
At the wide-angle position the trend is toward a barrel. For maxi- 
urn magnification the outer part of the field shows a pincushion, 
bile the inner part still produces barrel distortion. The amount of 
stortion shown in this diagram is only a fraction of 1 per cent. The 
i.phical presentation of lens aberrations has been greatly exagger- 
t d in order to show the way in which they change during the zoom. 
As it is impossible to enumerate all the different applications of the 
>omar lens to practical motion picture work, we can list only a few 

the more striking possibilities. 



Vol 49, No. 1 

FIG. 6. Ultra close-up of 
wrist watch. 

A combination of tilting, panning, and 
zooming is very difficult to achieve by con- 
ventional means, requiring perfect timing 
and the co-ordination of a highly skilled 
team of trained cameramen and techni- 
cians. Such a shot can be taken only in 
studios where cranes or booms with the 
necessary tracks, etc., are available. With 
the Zoomar, such a combination shot offers 
no difficulty at all and can be taken easily 
by one man without any assistance and 
without any costly implements. 

Educational and documentary films of- 
fer another wide field for the use of the 
Zoomar lens. Maps, pictures, and three- 
dimensional models can be filmed and their 
showing enlivened by zooming from over- 
all views to close-ups of details even if 
these pictures have to be taken in places 
where it is impossible to put up tracks for 
dolly shots. Thus filming with the 
Zoomar lens offers the possibility of intro- 
ducing movement and life into pictures of 
dead objects. 

The Zoomar lens should be very valu- 
able in the photography of sports events. 
On the race tracks, in the ball park, on 
the gridiron, to name only a few ex- 
amples, it would be very desirable if the 
cameraman could leave his box and shoot 
an interesting detail at close range. For 
obvious reasons, this is impossible. If he 
changes to a telephoto lens the exciting 
event is over when he is through with 
focusing his new lens, setting his dia- 
phragm, etc. The abrupt change of 
focal length together with the above- 
mentioned time lapse breaks up the 
continuity and makes it difficult for the 
spectator to understand the action on 

July 1947 ZOOMAR LENS 63 

the screen. With the Zoomar lens he can roam at will over the field 
and follow the horses or players wherever they go in one continuous, 
smooth shot and at the same time show all the details which con- 
tribute to the excitement of the event. 

Industrial shots can be enlivened and detail of machinery can be 
clearly shown with the Zoomar. Here again, the smooth and gradual 
transition from over-all shots to close-ups saves a lot of explanations 
which are necessary when long shots and detailed close-ups follow each 
other abruptly as has been customary. 

The same is true in the photography of commercial articles like 
costume jewelry. The zoom can be used to call attention of pros- 
pective buyers to details of workmanship and so on. 

In medical pictures, the use of zoom shots is nearly unlimited, 
especially in the field of surgery where close-ups of details are of the 
highest importance. At the same time, long shots have to be in- 
serted to facilitate orientation of the students who are going to see 
that picture. 

Even ultra-close-ups, like the inside of a wristwatch, can be pro- 
duced with a great amount of perfection (Fig. 6) . 

It is no exaggeration if we say that the close-up, this powerful means 
of expression given to the motion picture art by D. W. Griffith, has 
really come into its own by the introduction of the zoom shot. The 
Zoomar lens eliminates the difficulties which up to now have com- 
plicated the use of this technique. 


1 BACK, F. G.: "Zoom Lens for Motion Picture Cameras with Single Barre 
Linear Movement", /. Soc. Mot. Pict. Eng., 47, 6 (Dec. 1946), p. 464. 

2 BACK, F. G. : "A Positive Varifocal Viewfinder for Motion Picture Cameras", 
/. Soc. Mot. Pict. Eng., 45, 6 (Dec. 1945), p. 466. 



Summary. The equipment described was developed for analyzing underwatt 
motion of solid bodies. The experiments demanded a high rate of picture takin 
and that the subject studied should be in the field of at least two cameras at all time 
Edgerton-type flash lamps instead of shutter mechanisms were adapted, an endle. 
film belt giving a one-second exposure was developed, and a film speed of approx 
mately 35 ft per sec used. 

The Problem. For the past few years much of the work of th 
Hydrodynamics Laboratory of the California Institute of Tecl 
nology has been in connection with projects for Division 6 of th 
National Defense Research Council and for the Bureau of Ordnanc 
of the U. S. Navy. One of these projects required for its study th 
development of methods for making detailed measurements of th 
path and the orientation of a rapidly moving underwater body 
Consideration of the project showed that the work could be carrie< 
out in a horizontal, cylindrical tank about 30 ft long, with a depth o 
water of approximately 10 ft with a few feet of air space above th 
water surface. The desired range of experimental conditions im 
posed the necessity of varying the pressure from a positive value o 
two or three atmospheres down to a vacuum of a small fraction of ai 
atmosphere. This means that the tank had to be a closed pressur 
vessel and that very large windows would be both expensive an< 
hazardous. In spite of the difficulties involved, detailed study of th 
problem indicated that the photographic method of measuring th 
performance of the body was the most promising. 

Careful analysis of the probable experimental needs of the progran 
indicated that to secure the desired information, measuring point; 
would have to be obtained at a maximum rate of from 1000 to 300( 

* Presented Oct. 22, 1946, at the SMPE Convention in Hollywood. 
** Hydrodynamics Laboratory, California Institute of Technology, Pasadena 


per sec, depending upon the speed of the underwater body that was be- 
ng observed. However, since the size of the tank limits the length 
)f the path without regard to speed, the maximum number of points 
-equired for any one set of measurements was calculated to be about 
5000. If these measurements were to be obtained by the use of 
notion pictures, these requirements meant that camera equipment 
vould have to be developed that could take up to 3000 frames at a 
ipeed of from 1000 to 3000 per sec. Furthermore, an analysis of 
;he required accuracy of measurements showed that very little dis- 
;ortion could be tolerated. This indicated that the rotating prism- 
ype of high-speed camera would not be suitable. It was finally 
lecided that the best possibilities appeared to lie in the direction of 
L camera using a constantly moving film strip with no shutter mecha- 
lism, operating in conjunction with a battery of high-speed flash lamps 
yhich could act both as a source of illumination and as a shutter. 

The underwater bodies to be studied were of a type which would 
nter the water from the air at one end of the tank with a relatively 
dgh velocity. From the point of entrance on, they would be free 
odies and thus could move to any point in the tank, depending on 
he resultant forces that acted upon them. This meant that the 
ihotographic measuring technique had to be able to determine the 
osition of the body in space. In other words, it would be neces- 
ary to obtain the horizontal and vertical co-ordinates of the body 
n a plane normal to the axis of the recording camera, but it also 
rould be necessary to obtain the horizontal distance of the body from 
he camera. The best way to do this appeared to be by the use of 
he stereoscopic effect, which meant that for each point recorded, 
he body would have to be photographed by at least two cameras. 

To ensure this condition, a design specification was established that 
i the nominal plane of focus, the set of cameras should be so spaced 
tiat their fields of view would have a 60 per cent overlap. As a 
ssult of this, together with the other experimental requirements, 
tie recording equipment developed into a bank of five cameras, 
paced at 4 1 /2-ft centers and operated from a common drive shaft, 
'his gives a coverage over the entire experimental region of the tank 
'he central 20 ft of the 30-ft tank has a two-camera coverage over 
le effective width and depth, with single-camera coverage on the 
RTO 5-ft end sections. Fig. 1 (a) and (b) shows the diagrammatic 
rrangement of the cameras in the tank with their overlapping fields 
F view. 



Vol 49, No. 1 







FIG. 1. Diagrams showing arrangement of cameras and their overlapping fields 

of view. 


Detailed Requirements of Cameras and Light Sources. In the 
design and development of a camera system of the type just out- 
lined, it is necessary always to bear in mind the extreme importance 
of the relationship between the camera and the lights because the 
lights function as an essential part of the camera, since they replace 
the normal shutter mechanism. In fact, the characteristics of the 
flash lamps control much of the design of the camera itself. There- 
fore, these relationships will be discussed first in analyzing the de- 
tailed requirements of the over-all design. 

The most important characteristic of the flash lamp itself is prob- 
ably the effective duration of the flash. The minimum available 
flash duration controls the maximum usable film speed. It must be 
remembered that in this type of camera the film moves constantly. 
Therefore, the flash duration *of the lamp must be short enough to 
"stop" the motion of the film; otherwise, the recorded image will be 
blurred. One reasonable criterion of the maximum usable film speed 
is that the allowable amount of motion during one flash should not be 
greater than the circle of confusion of the lens of the camera. For 
most applications, this is a much more severe requirement than the 
one derived from a consideration of the possible image blur caused by 
the movement of the object being photographed. 

This difference can be seen easily by considering a set of typical 
conditions for the installation under discussion. The maximum 
speed anticipated for the body to be photographed is in the neigh- 
borhood of 200 to 300 ft per sec. Its average distance from the 
camera will be about 100 to 150 focal lengths. Thus 'the average 
speed of the image in the focal plane of the camera will be only about 
2*/2 ft per sec. Therefore, a film speed of 2 1 / 2 ft per sec would cause 
the same amount of blur for a given flash duration as would the move- 
ment of the object itself. It will be remembered, however, that 
experimental requirements call for a camera which :an take from 
1000 to 3000 frames per sec. Obviously, it would be impossible to 
take 3000 pictures per sec on a film having a speed of 2 l / 2 ft per sec 
since the frame height would be less than l / m of an inch. This means 
that a film speed of at least 10 times this value must be obtained. 
Obviously, any light which has a short enough flash duration to elimi- 
nate blur caused by film motion at this higher film speed will have no 
difficulty in stopping the motion of objects going at much higher 
speeds than those that will be encountered in this application. 

Investigation of the minimum .possible effective duration that could 

68 KNAPP Vol 49, No. 1 

be obtained from specially designed flash lamps indicated that 1 to 2 
microseconds was the shortest flash that could be anticipated. This 
duration permits a film speed of from 25 to 40 ft per sec to be used 
safely in the camera. On this basis the value of 35 ft per sec was 
selected as the design objective. 

It will be seen that there is no possibility of getting 3000 full-sized 
frames on a 35-ft strip of 35-mm film. In fact, the permissible frame 
height is only about 1 /& in. instead of 3 / 4 in. On the other hand, 
the dimensions of the tank and the possible camera installation in- 
dicated that it would be very desirable to have as large a frame height 
as possible in order to cover the working depth of the water. The 
obvious way to solve these conflicting requirements appeared to be to 
use one of the standard tricks long employed with flash-lamp il- 
lumination, i. e., to provide a black background and take multiple 
exposures. This is permissible in the present application since the 
speed of the body under study will never be high enough to cause 
images from successive flashes to fall on top of one another if the 
film moves as much as 1 /s in. between exposures. 

The second important characteristic of the flash lamp for such use 
is the intensity of the illumination. It is apparent that this intensity 
must be extremely high to produce an image of reasonable density in 
the very short flash time available. The magnitude of the problem 
can be seen more graphically if the operation of this type of equip- 
ment is compared to that of a hypothetical motion picture camera of 
the standard type using a normal shutter, but operating at a speed 
of 3000 pictures per sec. Such a camera would probably have a 
shutter opening of about 180 deg, which corresponds to an effective 
exposure time of Veooo of a second. Such a short exposure would cer- 
tainly require very-high-intensity illumination if a reasonably good 
negative were to be secured. However, Veooo of a second is a little 
over 160 microseconds. This means that, for a flash duration of 
one microsecond, the intensity must be at least 160 times as great as 
would have been necessary with a hypothetical camera using a 
normal shutter. 

The phrase "at least as great" was used advisedly, since there is 
some evidence to indicate that the reciprocity law breaks down for 
such extremely short exposures with the result that even higher- 
intensity illumination may be required to obtain a satisfactory image 
on the film. This particular installation imposes some extra demands 
on the illumination because the pictures must be taken under water 


and the long underwater light path means large light absorption. 
Taken together, these conditions all point to the fact that there is 
no possibility of securing sufficient illumination from a single lamp 
capable of flashing as rapidly as is necessary to meet our require- 
ments. Therefore, a synchronized battery of lamps is required, 
which, although it offers the possibility of solving the problem of suffi- 
cient intensity illumination, adds another difficulty of its own. If 
the flash duration is to be only 1 to 2 microseconds, then, if the 
lamp battery is to be effective, the lamps must be synchronized to 
flash simultaneously within a limit of about 1 /io of a microsecond. 

It may be of interest to consider the physical significance of the 
duration of the illumination of these flash lamps. The speed of light 
is approximately 186,000 miles per sec. Thus, in one microsecond, 
light can travel something less than 2 /io of a mile, or roughly 1000 ft. 
In other words a one-microsecond flash of light is only 1000 ft long, 
and if two such flashes are synchronized to Vio of a microsecond, they 
will run along neck and neck with their noses not over 100 ft apart. 

Lens Selection. As previously stated, the fact that the tank 
must withstand both pressure and vacuum precludes the use of 
large windows, hence, the cameras must be mounted very close to 
the tank. Likewise, the tank must be kept as small as possible to 
avoid undue expense in its construction and operation. This means 
that wide-angle lenses must be employed if the number of cameras 
required to cover the experimental area is to be kept within reason. 

At the time the design was started, lenses of one-inch focal length 
were the widest angle lenses available for the use of 35-mm film. 
Calculations showed that these lenses would cover the entire vertical 
field, assuming no refraction at the transition from air to water. 
However, when refraction was considered, the angle of view was 
reduced to the place that would necessitate two banks of cameras, 
one above the other, to cover the total depth of the water. It was 
obvious that two banks of cameras would introduce very serious com- 
plications, such as expense of construction, complication of operation 
and maintenance, and increased difficulty in analyzing results. 
Therefore, a more acceptable solution of the problem was sought. 
The obvious method of attack was to find some way of eliminating 
the reduction in the field caused by the refraction at the air-water 
interface, since without this refraction a single bank of cameras would 
be sufficient to do the job. 

A possible solution seemed to be the use of spherical windows with 

70 KNAPP Vol 49, No. ] 

which all the light rays will pass through the interface at an angle oi 
90 deg and therefore suffer no refraction. This possibility was re- 
ferred for analysis to Dr. Leonard M. Ross, consultant for Mt 
Wilson Observatory. 

A series of careful calculations showed that spherical windows 
were feasible and that the optical distortion would be no worse than 
in air if a satisfactory radius of curvature were employed and if the 
camera lens were mounted so that its front nodal point was at the 
center of curvature of the window. This combination results in a 
slight decrease in the field of view of the lens, since the addition oi 
the spherical window is equivalent to adding another element to the 
lens system. Its effect is to decrease the apparent distance between 
the original lens and the object. Thus, for example, if the distance 
from the lens to the object is actually 12 ft, the lens must be set foi 
a focal distance of about 2 ft when used with the spherical window 
For a one-inch lens, this decreases the field of view about 4 per cent, 
which is not very serious. 

Film-Magazine Requirements. The normal type of film maga- 
zine for a motion picture camera is, of course, one which uses twc 
spools, a supply spool for the unexposed film and a take-up spool foi 
the exposed film. However, this system is not well adapted tc 
film speeds as high as 35 ft per sec. To obtain such a speed starting 
with the film at rest requires a leader many feet long, even though 
the spools are accelerated as rapidly as is possible without film break- 
age. Decelerating the film at the end of the exposure is also a prob- 
lem if fraying of the film and the consequent filling of the camera 
with small film fragments is to be avoided. 

The installation contemplated here offers the unique feature that 
the cameras are looking into a completely dark tank until the flash 
illumination is started. This makes possible the use of an endless- 
belt type of magazine into which the required amount of film for one 
run can be loaded and cemented into a continuous strip. With such 
an arrangement the film can be run through the camera over and 
over again, thus making it possible to accelerate the film at a com- 
pletely safe rate until the desired speed is reached. This speed can 
be set exactly and held without variation during the time of the ex- 
posure, after which the film can be decelerated and brought to a stop 
with no danger of damage. Such a system appeared particularly 
desirable for use with a bank of cameras, since it would not only re- 
duce the film consumption, but also the danger of film breakage and 


similar troubles, and thus would increase the possibility of obtain- 
ing successful records from all cameras. 

The main requirements for the film drive have already been in 
dicated, i.e., smooth acceleration and deceleration to prevent dam- 
age to the film. However, the fact that this film was to be used for 
precise measurements made it desirable to maintain a constant film 
speed during the exposure and also a fixed relationship between the 
film speed and the number of flashes. This latter requirement is, of 
course, equivalent to the specification of a constant frame size, and, if 
it can be made to work satisfactorily, it offers a possibility for the 
development of a relatively simple projector to present the record 
in the form of ultra-slow-motion movement. 

Since these cameras are to be used as measuring instruments, they 
must be as carefully aligned and as permanently fixed to the windows 
of the tank as would be necessary for the use of any other optical 
measuring instrument. This means that the magazines must be 
of the daylight-loading type since the cameras cannot readily be 
removed from the tank and taken into the darkroom for reloading. 
Obviously, if a daylight-loading type of magazine is to be used, each 
strip of film will contain a small exposed portion which can also be 
made to contain the splice. To obtain a complete record of the 
experiment, measurements must be started at the beginning of the 
entire unexposed length of film, or, in other words, just after the 
exposed portion, including the splice, has passed through the camera. 
If this condition is to hold simultaneously for all of the cameras in 
the battery, then the lengths of the film in all of the magazines must 
be identical ; i.e., they must have the same number of sprocket holes 
per film belt. If this condition can be otained, then all of the splices 
in the exposed portions of the film can be set at the same relative 
position and will maintain this relation during the entire run. 

To secure identical lengths of film in each magazine requires a 
precision-loading technique, a prerequisite of which is an exact means 
for measuring the film being loaded. Furthermore, an endless-belt 
type of magazine demands a fixed predetermined pattern of threading 
which must be maintained during the loading and unloading cycles. 
The best way of meeting these rather complicated demands appears 
to be through the construction of a magazine loader which would 
incorporate a film-measuring device, a splicer, and film supply and 
take-up rolls of sufficient size to load the entire camera battery a 
number of times. Since this auxiliary piece of equipment is necessary 



Vol 49, No. 1 

for the successful operation of the camera battery, it can be con- 
sidered as a necessary part of the camera equipment, and therefore 
a description of it will be included. 

Description of the Camera. The camera itself is basically a very 
simple device. Inherently, it has a fixed focus since with the lens 
wide open the depth of focus is sufficient to cover the entire operat- 
ing portion of the tank. The camera consists essentially of two 
elements, the lens and its mount, including the spherical window, 
and the film drive which incorporates the special gate at the focal 
plane. Fig. 2 shows a sectional elevation of the camera itself. 


FIG. 2. Sectional elevation of high-speed camera. 

- The lens system can be seen in this figure. It will be observed 
that the lens is mounted in a barrel having a fine pitch thread for 
initial focal adjustment. The mounting for the spherical window has 
been given a great deal of consideration. It should be remembered 
that water is always in direct contact with the convex side of this 
window. The concave side, as well as the camera, is in air and re- 
mains at atmospheric pressure. On the convex side the water pres- 
sure may vary from about 14 Ib per sq in. below atmospheric pressure 
to 45 Ib per sq in. above it. Since this spherical window is actually an 

July 1947 



auxiliary lens for the camera, precise alignment must be maintained 
at all times. This is ensured by installing the glass in direct contact 
with a metal mounting flange. Leakage is prevented by the use of a 
rubber-ring gasket of the unsupported-area type which works equally 
well for either positive or negative pressures. Fig. 3 shows the ap- 
pearance of this spherical window as seen from the inside of the tank. 
The camera lens is visible in the center of the window. 

The film path through the camera is very simple, as may be seen 
in Fig. 2. It enters the camera near the top, passes over a roller 
which guides it into the focal plane, goes down through the gate to 
another guide roller, which feeds it to the drive sprocket. The film 
has a 90-deg wrap on the drive sprocket and leaves it with a correct 

FIG. 3. Spherical window as seen from 
inside of tank. 

FIG. 4. Detail of roller guides for focal 

alignment to go directly to the first idler spool in the magazine. The 
gate itself is of special design to operate at the highest film speed that 
is used. It was believed that it would be extremely difficult, if not 
impossible, to prevent film damage if rubbing contact were permitted 
at any point in the camera with a film speed of 35 ft per sec. At the 
same time, exact positioning of the film in the focal plane was required 
in order to secure quantitative measurements from the records. 

The system of roller guides, seen in Fig. 2 and shown in detail in 
Fig. 4, was developed for this purpose. It will be noted that the 
rollers are relieved between the sprocket holes so that they do not 
touch the film in the active area. All of the rollers in the camera 
and magazine are so relieved. Edge rollers are provided to define 



Vol 49, No. 1 

precisely the lateral position of the film. A double pair of rollers at 
entrance and exit was found necessary to flatten the film properly as it 
passes through the focal plane. Tests showed, however, that this 
construction alone was not sufficient to ensure complete flatness of the 
film under all conditions since a slight lateral curvature tended to 
persist. The only successful method yet developed of eliminating 
this curvature has been through a rough control of the humidity by 
use of humidifying pads in the magazine. Fig. 5 is a side view of the 
camera showing the loading door. The opening and closing of this 
door also moves interlocking pins which operate the light locks so as 
to prevent accidental exposure of the film. 

Fig. 6 shows the external view of the film magazine. It will be 

noted that it seems dispropor- 
tionately large compared to the 
size of the camera. This is, of 
course, because of the fact that 
the film is laced as an endless 
belt in the magazine. Fig. 7 (a) 
and (b) shows the external con- 
struction of the film rack with 
the film laced in place. The run 
on which the film enters and leaves 
the camera is a straight vertical 
run, whereas most of the others 
are diagonal. The alternate use 
of large and small spools on 
both the upper and lower shafts 
eliminates the possibility of the 
film's touching and rubbing on 
the crossover points. The shaft 
which carries the lower set of 
spools is loaded both by gravity 
and by an auxiliary spring in order 
to secure the desired film tension. 
The film lacing shown in Fig. 7 
utilizes the maximum capacity 
of the magazine. This lacing 
can be modified to accommodate 
shorter lengths in case a full- 
FIG. 6. External view of film magazine, length film strip is not necessary. 

FIG. 5. 

Side view of camera showing 
loading door. 

July 1947 



FIG. 7. External construction of film rack with film laced in place. 
(a) Front. (b) Back. 

Camera Drive. Fig. 8 is a general view of the bank of five 
cameras mounted on the side of the tank. The common drive 
shaft can be seen passing through all of the cameras. This drive 
shaft was designed for high torsional rigidity to eliminate any ap- 
preciable displacement angle between the drive sprockets of the 
different cameras. To prevent any binding at the bearings, flexible 
universal joints of the metallic-disk type have been provided at 
each side of each camera. When the battery is being operated, with 
full magazines running at the maximum speed of 35 ft per sec, the 
angular displacement of the drive sprockets between the first and last 
cameras in the bank is less than 0.04 deg, which is equivalent to 0.001 
in. of film travel. For most conditions of operation, this introduces 
an error in the measurements too 
small to be significant. However, 
it can be corrected very easily by 
adjusting the corresponding pro- 
jectors to have the same angular 
displacement with respect to the 
first camera in the bank. The syn- 
chronous drive motor can be seen 
on the extreme left of the drive 
shaft. Fig.Qisamoredetailedview FIG. 9. Synchronous drive motor. 

FIG. 8. Bank of five cameras mounted on side of tank. 

76 KNAPP Vol 49, No. l 

of this motor. It will be seen that there is an electric brake on each 
end of it. These brakes are used to secure the low and smooth accel- 
erations and decelerations, required to prevent film damage. It will 
be observed in Fig. 9 that the entire motor is mounted on trunnion 
bearings so that it can rotate. The power supply is brought to it 
through a set of three slip rings mounted on the outside of the motor 
frame. The electric brake on the right-hand end operates on the 
motor shaft ; whereas the one on the left-hand end works against the 
motor frame. The operation of this system can best be seen by fol- 
lowing through a complete cycle, as follows : 

Consider that the cameras are all loaded with full magazines and 
that everything is stationary but ready for a normal photographic 
run. The motor-shaft brake is then clamped and the motor-frame 
brake released. The starting switch is then closed, applying power to 
the motor. Since the rotor and drive shaft are clamped, the motor 
frame starts to revolve on its trunnion bearings. It accelerates up to 
normal speed and synchronizes. Next the shaft brake is released, 
making it possible for the rotor and drive shaft to turn. The frame 
brake is then applied gradually at a predetermined rate. This slowly 
brings the motor frame to a stop while the rotor with the drive shaft, 
camera sprockets, and film magazines all come up to full synchronous 
speed. The cameras are now ready to record the results of the experi- 
ment. This is done by launching the body, the motion of which is to 
be recorded, and simultaneously turning on the battery of flash lamps. 

When the exposure is made and the record obtained, the film is 
brought to a gradual stop by a reversal of the above procedure; i.e., 
first the frame brake is released so that the frame is free to turn, then 
the shaft brake is applied at a gradual rate, and the film in all 
the cameras brought to a standstill. By means of a series of inter- 
locking relays, all of these operations are performed automatically 
after the starting or stopping switch is closed by the operator. 
" One further refinement is incorporated in the procedure. The cam- 
era drive shaft is connected to a simple gear train which acts as a 
counting device to keep track of the location of the film splices. This 
counter is interlocked with the main control of the experiment itself 
so that the experiment can be started and the flash lamps energized 
only immediately after the film splice and the exposed section of the 
film have gone through the camera. This interlock ensures that 
full record of the phenomenon under investigation will always be 
obtained at each run. 


Magazine Loader. The film magazines are all detachable from 
the cameras and may be taken to the darkroom for loading. How- 
ever, they are large and bulky and quite heavy, even though made 
of aluminum. As explained previously, a special loading device is 
necessary in order to preserve the threading pattern and also to en- 
sure that each magazine is loaded with exactly the same length of 
film. A little study showed that it would be simpler to design the 
magazine loader so that it could be brought to the magazine, rather 
than to take the magazines into the darkroom for reloading. To 
facilitate this operation, a special bracket was installed at each cam- 
era to permit the magazine to swing down in a horizonal position 

FIG. 10. Magazine in loading position. 

behind the camera. Fig. 10 shows one of the magazines in this 
loading position with the loader mounted in place on it. A standard 
Mitchell 1000-ft camera magazine is used for the unexposed and ex- 
posed film storage. The lower section of the loader contains the 
sprocket-hole counter and the splicer. 

The operation of the loader is as follows : It is first placed upon the 
magazine using the same locating dowels that position the magazine 
on the camera. The loading-film loop containing the splice projects 
out of the magazine into the loader. This loop is now broken. One 
end of it is spliced to the end of the exposed film going to the take-up 
spool. The other end is spliced to the unexposed film coming from 
the supply spool. The loader door is now closed, which opens the 

78 KNAP-P Vol 49, No. l 

light locks to the magazine and to the supply and take-up spools. 
The operating crank is then turned until the exposed film has been 
drawn out of the magazine and a reload of unexposed film drawn in. 
The correct amount of film is indicated by the reading of the sprocket- 
hole counter. The loader door is then unlocked, which closes the 
light locks into the camera and film magazine. The film is then cut 
from the feed and take-up spools and a splice made to form the end- 
less belt in the camera magazine. This splice is made at the exact 
sprocket hole indicated by the counter. It will be seen that all of 
the splicing and cutting can be done in daylight since the only film 
exposed is that section which it is necessary to use in any case during 
the threading of the loop through the camera gate and drive 

Flash Lamps. The flash lamps are straight gas-filled tubes 
approximately 8 in. long and are made of quartz. Quartz tubes 
are necessary because of the tremendous amount of energy that is 
fed into them to produce the flashes. The use of quartz makes 


FIG. 11. Flash lamp. 

possible much longer runs than could be obtained by Pyrex or any 
other type of glass tube. The characteristics of these tubes limit the 
length of record that can be secured when being operated at flash 
speeds of 1000 per sec or greater. Their effective limit is about 3000 
flashes. At this point they become so hot that the duration and in- 
tensity of the flash is seriously 
affected. Continued operation 
beyond 3000 flashes quickly re- 
sults in softening and final col- 
lapse of the tube. Fig. 11 shows 
one of these tubes. 

Fig. 12 is a time-intensity 
record of the flash produced, 
FIG. 12. Time-intensity record of flash This record was obtained by 

Horizontal tin r e: d 2 divisions per bser g * ^ with a P hot - 
microsecond. cell driving a synchroscope. It 

July 1947 



will be seen that the duration at peak intensity is approximately 
one microsecond and the effective photographic duration between 
1 and 2 microseconds. The energy input to the tube is approxi- 
mately one joule per flash. Expressed in other units, this means 
that at a flashing rate of 3000 per sec, each tube requires an 
amount of energy equivalent to a continuous input of 3 kw. If the 
duration of each flash is assumed to be one microsecond, the light is 
on Vs33 an d off 332 /333 of the total time. Thus the rate of energy input 
to one tube during each flash is about one megawatt (1000 kw). 

The flash tube is made in the 
form of a line source because the 
area to be illuminated is rec- 
tangular. To utilize the light 
as efficiently as possible, each 
tube is mounted in a special re- 
flector as shown in Fig. 13. 
These reflectors are constructed 
of Lucite, cemented together and 
then aluminized by the vacuum- 
sputtering technique. The cross 
section of the reflector is so de- 
signed that the light is dis- 
tributed as uniformly as possible 

over the depth of working section in the water tank. These lamps 
are normally used in banks, jointed together as shown in Fig. 14. 
This construction makes it possible to insert them into Lucite tubes 
of about 8 in. in diameter. These tubes run longitudinally through 
the tank and are provided with stuffing boxes at each end to pre- 
vent leakage. Thus, the lights themselves are in air at atmospheric 
pressure at all times, but so far as illuminating the tank is concerned, 
they act as if they were under water. 

A detailed description of the power supply and control circuits for 
these lamps is beyond the scope of the present paper. They are 
basically the product of Prof. Harold E. Edgerton and his assistants 

FIG. 13. 

Flash tube mounted in 

FIG. 14. Bank of high-speed flash lamps. 



Vol 49, No. 1 

at Massachusetts Institute of Tech- 
nology, who are pioneers of many 
years standing in this field. Rather 
extensive modifications of the normal 
equipment were incorporated in this 
development for the Hydrodynamics 
Laboratory, which shortened consider- 
ably the duration of the flash and also 
made it possible to operate the lamps 
with cables of varying lengths up to 
100 ft connecting them to their re- 
spective control units. Measurements 
taken during operation indicate that 
the circuits employed in the control 
units showed deviations in flash time 
of 0.1 microsecond or less. This 
means that the flash duration of the 
entire battery of lamps was not over 
10 per cent longer than that of any 
individual tube. Each lamp requires 
a separate control circuit. Fig. 15 
shows a battery of six such units 
with the cabinet door open to give 
a view of the interior. In all, seven 
batteries have been installed, giving a 
maximum of 42 lamps that can be 
operated in synchronism. 

The fundamental principle of opera- 
tion is very simple. Each unit con- 
tains a high-voltage capacitor, which 
acts as a power reservoir to supply 
the necessary energy to operate the 
light during the flash period. Power 
is poured into this capacitor during 
the relatively long period between 
flashes. At the time of the flash, the 
entire amount of energy stored in 
the capacitor is discharged through 

FIG. 15. Battery of six control the lamp tube through a high-capacity 

Thyratrpn tube, which is used for the 



control switch. The power source for charging all the capacitors of 
the lamp battery is a 120-kw three-phase, high-voltage rectifier 
which supplies d-c power at any desired voltage up to 6000. The 
control circuits incorporate voltage doublers, thus the lights them- 
selves can be operated at any desired potential up to 12,000 v, de- 
pending on the light intensity desired. Needless to say, circuits 
operating with d-c voltages and capacitors of these magnitudes re- 
quire safety precautions for the operating personnel. 

Operating Experience. The cameras that have been described 
and their associated equipment are now coming into full operation. 
It is felt that the installation will prove to be a very keen tool for 
the study of high-speed hydrodynamic phenomena. It may be 
of interest to know that one of the major difficulties that has had 
to be overcome in obtaining satisfactory results with this equipment 
has been the securing and maintaining of a supply of water having 
satisfactory clarity. It will be remembered that the length of the 
light path from the lamps to the object under study and back to the 
film in the camera is about 25 feet. 

Measurements made with dust-free double-distilled water, purified 
with the utmost care, indicates that a layer of such water one foot 
thick will transmit 99 per cent of light that falls on its face. At first 
this sounds well, but it must be remembered that the light path is 
25 ft long. To calculate the light transmitted through 25 ft of water, 
it is necessary to raise the coefficient of light transmission to the 25th 
power: 0.99 to the 25th power is 0.78. In other words, one fourth 
of the light will be absorbed and three fourths will reach the camera. 
The corresponding absorption for a 25-ft air path is immeasurably 
small. Thus it will be seen that the very best water obtainable ab- 
sorbs an appreciable amount of light. 

However, the Laboratory has made comparative measurements 
of a number of different samples of water, all of which appeared very 
clear to the eye when viewed in 5-gal containers. It was found that 
carefully filtered de-aerated city water stored in glass until the time of 
measurement had a transmission coefficient of 0.95. Identical water 
stored in a rubber-lined tank for two weeks showed a transmission 
coefficient of about 0.65. Another of the same samples stored for 
the same length of time in a tank painted with high-grade zinc 
chromate showed a transmission coefficient of about 0.85, but, in 
addition, it was found that the transmission in the blue was com- 
pletely eliminated. 

82 WARREN Vol 49, No. 1 

Table 1 shows a comparison of these various water samples on the 
basis of a 25-ft light path. An examination of this table makes only 
too evident the absolute necessity of securing a clear water supply 
for underwater photography which requires an appreciable light 
path. A major step in the solution of the problem for the Hydro- 
dynamics Laboratory tank has been to line the tank completely 
with Koroseal of a type that was first subjected to rigorous tests to 
prove that it did not affect the water stored in it in any way. 



Transmission of Light 

Coefficients Transmitted per 

Water Sample per Foot 25-Ft Path 

Filtered distilled 0.99 78 

Filtered city supply . 95 28 

Zinc chromate paint container . 85 2 

Rubber-lined container 0.70 0.01 


Summary. This paper deals with future AAF requirements for motion picture 
equipment to meet such extremes as exposures at altitudes of 100,000 ft and speeds 
of 1500 mph. It discusses three wartime cameras which are focused on things to 

Approximately eleven years ago, Colonel Goddard of Air Materiel 
Command addressed the Society of Motion Picture Engineers. At 
that time, looking forward to the then spectacular future, he prophe- 
sied photographic aircraft traveling at the rate of 300 mph. That 
was eleveil years ago. Hiroshima and Nagasaki were but names on a 
map. Operation Crossroads and the White Sands guided-missile 
projects were unknown. Today, while we talk about aircraft travel- 
ing at the rate of 1500 mph at altitudes of 100,000 ft, there are 
future projects and operations of which we can only guess. One fact, 
however, which was true eleven years ago is still true today the 
Air Corps must and will be ready to photograph the things to come. 

* Presented Oct. 25, 1946, at the SMPE Convention in Hollywood. 
** Air Materiel Command, Wright Field, Dayton, Ohio. 


While tomorrow's photographic requirements are still subject to 
conjecture, there are some clear signposts. Aerial cameras must 
operate from a few feet off the ground to miles above it. They must 
literally be able to stop the effects of supersonic speeds ; and finally, 
they must be able to be airborne whether by aircraft, rocket, or pro- 
jectile. The major problems are height, speed, power, space, and 
weight. Let me reverse the problems, and touch on the last ones 
first. Throughout the war we experimented with longer and longer 
focal lengths, which has meant consistently larger and heavier 
cameras. We are now testing a 100-in., //10, 600-lb camera, and 
in our agenda is a 240-in. camera. Obviously there has to be a limit. 
Even though that limit may be far removed, it is important that we 
realize the time is coming when cameras borne in aircraft may be 
outmoded and aerial photographers out of jobs. Already, remotely 
controlled 35-mm turret-installed cameras have been used successfully 
at Bikini, in the drone flight from Hawaii, and in the White Sands 
project. In the future, with speeds beyond sound, the design, cubage, 
and weight of photographic equipment will be increasingly important. 

Our next problem is to capture clear pictures at rocket and jet 
speeds. Cameras with increased focal lengths comprise one method 
we are using. We are also experimenting with the possibility of a 
variable-opening shutter with maximum lens opening so that we may 
obtain the highest possible shutter speed for varying light conditions. 
We already have under procurement a high-cycling strike camera 
with a shutter speed of 1 / 2 ooo of a second. In the future we shall be 
experimenting more and more with electronic means for taking pic- 
tures at high speeds. Experiments are being conducted on films 
having deep red sensitive emulsions in order to attain greater film 
speeds. At Bikini, our cameras were required to photograph every 
millisecond of the atomic bomb blasts. 

Man's continual climb toward stratospheric heights has presented 
us with other problems, and not the least of these is temperature 
control at high altitudes. We have made use of thermostatically 
controlled heating elements and automatic air-density compensating 

The Baker 40-in. telephoto lens is a successful solution to this 
problem. The rear lens of this camera floats on a link system of 
small bellows which expands and contracts to change the focus auto- 
matically in order to compensate for air pressure and temperature 

84 WARREN Vol 49, No. 1 

Among the other problems arising from higher and higher altitudes 
are corrections for increased aerial haze, vibration, and the manu- 
facture of lenses which will permit pictures of vast areas. In the use 
of color film, we are working toward films with higher speeds and 
greater contrast particularly in the blue layer so that we may use 
sharper cutoff filters to reduce the loss of contrast caused by aerial 
haze. Pictures taken in jet-propelled aircraft have opened up a 
new era of aerial photographs having little or no vibration. 

In addition, we are continuing to work for a stabilized mount which 
will ensure absolutely vertical, vibration-free exposures. In the 
field of wide-angle cameras, one of the most unusual is the Baker 
Spherical Shell model which permits the taking of photographs that 
cover an angle of 120 deg, so that when used at a 30,000-f t altitude an 
area of 300 sq miles is obtained in each exposure. The experimental 
model, which only recently has been taken out of the secret class, 
makes use of a 4-in.,//2.8 lens that produces the images on a Spherical 
Shell. The negatives are projected in a special projection printer 
onto a white acetate base sensitized paper 40 X 40 in. square. Our 
experiments with guided missiles are providing us with records and 
information about those regions to which no man has yet ascended. 
Fortunately, they show that we are on the right track in our high- 
altitude research. 

In some cases we have more than signposts we have the pre- 
liminary models for some of tomorrow's cameras. Specifically, I 
should like to discuss three models which I believe qualify for the 
future. They are the 0-5^4 Radar Recording Camera, the A-6 
Motion Picture Camera, and the Sonne S-7 Continuous Strip Camera. 

My reasons for including the O-5A camera are many. First, it 
has already been used to obtain valuable atomic bomb data, photo- 
graph the pip of the moon, and record the flight of drone planes. 
Today it is being used in the White Sands project, and to photograph 
areas covered by clouds, which cannot be photographed with aerial 

Turning fr6m the present of the 0-5A camera for a moment, let 
us survey its history. Despite enormous strides and even more 
spectacular press releases, aviation as late as 1945 was -largely a fair- 
weather weapon and radar navigation left something to be desired. 
One has only to recall the Battle of the Bulge for substantiating the 
fact that the entire allied air force was helplessly pinned to the ground 
by weather while the German artillery had its field day. It was not 


until research pointed out that the fault lay not with the aircraft or 
radar, but with radar interpretation, that real progress was made. 
You have seen radar maps of cities, coastlines, etc., and probably 
have wondered how anyone could make sense out of the things 
let alone distinguish an island from a battleship. Imagine what it 
was like when there were no maps ! 

The 0-5 A camera did not just grow. Many trials had to be made 
and some errors committed before it first saw the light of radar. 
However, from the first radar recording camera, which appeared in 
1942 (a Kodak 35 with a total of 36 exposures), to the present model, 
much progress has been made. Today's 35-mm 0-5.4, the joint 
result of research by the Army, Navy, and MIT, has a 100-ft film 
magazine which can be replaced in 15 sec, giving thereby an 
almost unlimited number of exposures. No longer does the operator 
have to choose which of a few sweeps he can record. A simple push 
button permits him to set the camera for every sweep, every other 
sweep, every tenth, or every sixtieth sweep. No longer does the 
radar operator have to focus the main scope and then focus the aux- 
iliary scope used for the camera. An elbow beam splitter with a 
color-selective coated surface permits the same scope to be used by 
operator and camera. Throughout the development of the radar 
recording camera, we have consistently reduced its size and weight. 
The 0-5 A model with all its accessories weighs only 25 Ib and requires 
only 5 / 8 cu ft of space. 

Despite this progress, tomorrow's radar recording camera will have 
to be smaller and lighter, have an even larger film capacity, and be 
adapted to remote control perhaps from miles away. It will be 
used to chart the route for tomorrow's navigation, and provide the 
road maps for three-dimensional color photography 

Among general utility motion picture cameras, my nomination for 
the future is the A-6 model. Because AAF motion picture cameras 
in operation throughout the war did not satisfactorily fulfill all 
aerial requirements, the A-6 Model finally was designed. It is a 35- 
mm camera with operating speeds of 16, 24, 32, and 48 frames per 
sec and has an 87rdeg shutter which permits Vioo of a second exposure 
when operating at the rate of 24 frames per sec. 

Unlike many of the earlier models which were modifications of 
existing ground cameras, the A-6 was built primarily for aerial use. 
As a consequence, special features were required. Among these are a 
3-lens turret, which prevents vignetting with lenses between 25 mm 

86 WARREN Vol 49, No. l 

and 6 in., and a critical focusing attachment that can be inserted in 
the camera directly behind the lens in the "take" position to provide 
a means for accurately checking the field of view and for bore-sighting 
the camera. This camera has the ability to operate at temperatures 
as low as 65 F. Since operators working in low temperatures 
require gloves, which are often heavy and cumbersome, a special 
magazine was developed to permit easy loading of the camera. In 
addition, special start and stop buttons were designed for gloved 
operation. The buttons are placed 2*/2 in. apart, and the stop but- 
ton remains in the lock position until it is depressed. Taking into 
consideration the difficulties of aircraft installation, the A-6 is 
designed to operate in any position in regard to horizontal and 
vertical axes, and a direct image view-finder permits the field to be 
viewed as far as 12 in. behind the view-finder. To provide further 
utility, the A-6 may be operated by a detachable motor, a spring 
motor, or by means of a hand crank. An adapter is provided for 
tripod use. 

Although the performance record of this camera has yet to be 
written, plans for future development are already under way. Our 
main objective is to provide a motion picture camera with the greatest 
possible film capacity and focal length, and at the same time, have it 
require the smallest possible amount of space. Along this line we 
are working on 16- and 35-mm combat recording cameras which 
will provide speeds of 48 to 96 frames per sec, operate in any position, 
and have erector systems which will permit normally viewed pro- 
jected images regardless of the taking positions. Another motion 
picture camera on the docket is one having an automatic diaphragm 
with a photoelectric cell that operates from reflected light to adjust a 
variable-opening shutter. Development is also under way for a camera 
mount which will permit varied installation and, by means of servo- 
motors, permit the camera to be sighted by an automatic computer. 

Perhaps the most spectacular camera, however, is the S-7 Sonne 
continuous-strip stereoscopic camera. While it is not a motion 
picture camera in the true sense of the word, it does have a moving 
picture film feature. The most revolutionary feature about the 
camera is its lack of shutter. The film moves at a controlled speed 
past an adjustable slit permitting the desired exposure and at the 
same time compensating for ground motion. Another advantage is 
that it provides an exposure 200 ft long by 9V2 in. wide without 
breaks in the continuity. Because of the ground-image compensation 


and the continuous-strip features, this camera has been ideally 
suited to low-altitude, high-speed photography. Its history is filled 
with spectacular triumphs, of which I shall mention but a few. It 
was used to obtain color stereoscopic pictures of the beaches of Oki- 
nawa and the resulting exposures were used to chart the invasion. 
When the landings had been made successfully, the heights of the 
underwater reefs and sea walls obtained from the pictures were ac- 
curate to fractions of a foot. Beside making history in the Pacific, 
this camera was used to prepare the grounds for the Normandy in- 
vasion. At Bikini, with several modifications, it provided facts 
about nuclear science which would have been unattainable otherwise. 

The value of pictures taken at low altitudes and high speeds with 
stereoscopic features makes this camera of great value to the future. 
Today, despite the tremendous amount of photographic reconnais- 
sance accomplished, only one fifth of the world has been charted 
and that, for the most part, in black and white. In order to pinpoint 
projectiles, the first requirement will be for precise and accurate 
pictures taken only a few feet off the ground. Color pictures will 
divulge the product of a manufacturing plant by its smoke, residue, 
or raw products, and stereo pictures will reveal accurate heights and 
depths. These are necessities for tomorrow. 

Already we have flown the S-7 Sonne Strip Camera at speeds of 
580 ft per sec, only a few feet above the tree tops, and have made 
plainly visible nails in the planks of bridges and stickers on the wind- 
shields of cars. We must be able to do the same thing traveling at 
speeds faster than that of sound. 

Despite its satisfactory record, the strip camera is slated for future 
modification. Instead of interchangeable cones of a single 6-in. lens 
dual 88-mm, 100-mm, or 5-in. lenses, stereo cones accommodating 
two 12-in. and two 20-in. lenses are under procurement. Designs 
have been drawn up for a nonbanding precision film-drive mechanism, 
a stabilized mount, and a film capacity of 400 ft. Under develop- 
ment is a new type stereo-strip camera which employs a single 40- 
in. f/5 telephoto lens. 

This camera will have two slits to provide stereo photographs. 
The present system uses two lenses located fore and aft in respect to a 
single slit. Two separate rolls of 400-ft film will be used. It is ex- 
pected that this camera will permit extreme high-altitude photog- 
raphy in jet-propelled aircraft. 

Let me at this point repeat a statement I made earlier: we are 

88 LOCKNER Vol 49, No. 1 

proud of these cameras, and the entire photographic industry may 
justifiably be proud of its record. However, these cameras of which 
we are so proud are simply crude, experimental models of tomorrow's 
photographic equipment. The past and the present belong to his- 
torians. The future is our concern. Years of hard, and often dis- 
couraging, research have brought photography into its own. There 
is no other way, no primrose path, into the future. Research and 
research alone will get our cameras into the stratosphere to road-map 
the future at supersonic speeds. 


Summary. Wartime requirements of airborne armament led to the development 
of a successful system of stabilization which, when applied to gun turrets, enabled 
gunners to track a target with greater ease. At present, the system is being adapted 
to the stabilization of aerial cameras to overcome certain faults inherent in present 
photographic techniques. This paper discusses the problems and describes the sys- 
tem applied to aerial photography. 

The techniques of motion picture photography and aerial photog- 
raphy appear to have at least one problem in common. Any motion 
of the studio camera mount during the interval of making a shot will 
produce undersirable results in the final pictures. In aerial photog- 
raphy any deviation of the airplane from smooth flight will cause 
tilt of the photographs and also blurring of the photographs resulting 
from image motion during exposure. It has been generally recognized 
that some form of stabilization of the camera mount would alleviate 
this condition, if not provide a satisfactory and complete solution. 

The problem of developing such a system of stabilization has re- 
ceived the attention of many talented organizations over a long period 
of time. To date, no practical stabilizer has been evolved which gives 
any measure of satisfaction to the critical demands of the photographic 
engineers. Our own engineers have directed a great amount of 
energy in this direction at various times over the past twenty years 

* Presented Oct. 22, 1946, at the SMPE Convention in Hollywood. 
** Fairchild Camera and Instrument Corporation, Jamaica, N. Y. 


with no apparent success. However, the requirements of wartime 
aircraft armament development received a great amount of emphasis 
in our engineering laboratories. During the course of the many devel- 
opments completed in that field, a successful system of stabilizing 
aircraft gun turrets was evolved. It appears that it may be possible 
and practicable to modify this stabilization system to provide a fairly 
satisfactory means of stabilizing a camera mount. 

Early gunsight computers used in aerial gun turrets had to start 
from many assumptions. The greatest assumption was that the air- 
craft always should fly straight and level. Any deviations from such 
straight and level flight would mean that the gunner would have to be 
adjusting his turret constantly to take care of any aircraft maneuvers. 
Such variations from straight and level flight caused false information 
to be fed to the computers, and caused constant confusion to the gun- 
ner, since he never could predict completely the direction and magni- 
tude of any such deviation from prescribed conditions. 

At the same time, the bomber, when participating in a dog fight, 
was literally a sitting-duck target for any enemy aircraft. The enemy 
knew very well that once the bomber was to take a certain course, it 
could not deviate from that course without upsetting the accuracy of 
all the armament installed in the bomber. This vulnerability made 
it very desirable that some new instrument be developed which would 
enable the bomber to fly a devious and eccentric course in order to 
avoid the hazards of being a so-called sitting-duck target without at 
the same time compromising the accuracy of its armament. 

Consequently, early in the war, a very well-defined requirement 
arose for some system of stabilization to be developed so that a turret 
installation in a bomber or fighter would remain in a fixed position in 
space, irrespective of any maneuvering tactics the carrying aircraft 
might assume. 

This problem of supplying stabilization was approached from many 
angles. All solutions had one thing in common a means of over- 
coming the primary control of the power system of the turret to cor- 
rect its motion in reference to the airplane so that its control might be 
established in terms of outside space. Such control in terms of out- 
side space is predicated on some means of measuring position or rate 
of the turret in outside space. Of the many means possible, the two 
most commonly used are a vertical gyroscope and a rate gyroscope. 

The Fairchild stabilization system employs a rate-measuring gyro- 
scope as a control instrument. Two gyroscopes are used and are so 

90 LOCKNER Vol 49, No. 1 

oriented that rates are measured in the horizontal and elevation 
planes. These two instruments are identical and interchangeable. 

The gyroscope rotor is driven by a speed-controlled electric motor 
and revolves at 12,600 rpm. This rotor is mounted in gimbals and is 
restrained by calibrated springs with a single degree (or plane) of free- 
dom permitted. An angular motion perpendicular to this plane of 
freedom results in a gyroscope precession directly proportional to the 
applied angular velocity. When such gyroscope precession occurs, 
the magnetic circuit of a balanced E pickoff becomes asymetrical, and 
an ouput voltage is developed. Any angular rate up to 33 deg per 
sec has a corresponding gyroscope output voltage. Moreover, the 
direction of the angular motion is determined by the phase relation- 
ship of the output signal. 

By maintaining this gyroscope output at a given level, any desired 
rate in space can be established. Should any deviation from the 
desired rate occur, an immediate change in gyroscope output is 

The method used to select and maintain this signal level is as 
follows : 

The gyroscope E pickoff is supplied by a 1 10-v, 400-cycle source of 
power, usually the aircraft inverter. This same source of power is 
used to supply a transformer delivering its secondary voltage to a 
potentiometer geared to the hand controls of the turret. Any rotation 
of the hand controls to require a velocity from the turret rotates this 
potentiometer. This rotation results in a voltage output increasing 
from zero at the center of travel. Equal displacements right and left 
of the center produce voltages equal in magnitude but opposite in 

For a given rotation of the hand controls, a given voltage is pro- 
duced. The difference between this voltage and the gyroscope out- 
put is fed into an integrating servo whose output turns another po- 
tentiometer. This servo-controlled potentiometer, through an ampli- 
fier, controls an amplidyne generator, and finally, the turret rate. 
The system compares the gyroscope output with the chosen output of 
the hand-control potentiometer which represents the desired rate in 
space. If any difference exists, the servopotentiometer rotates, 
changing the turret rate until the gyroscope output just equals that 
of the hand-control potentiometer. Any change of angular rate of 
the turret from an external cause, i. e., rough air or turning of the air- 
craft, will result in a changed signal from the gyroscope. Its voltage 


will no longer be equal to that of the hand-control -potentiometer 
and an appropriate correction in turret rate will be initiated through 
the servo-controlled potentiometer. 

Quantitative measurements made on this system indicate that the 
activating or drive mechanism of the turret presents the most serious 
drawback to a good stabilizing system. A drive system that is quick 
to respond to changes in an input signal will accumulate a much 
smaller error in velocity or acceleration than a system with sluggish 
response. The amplidyne generator drive used in the system under 
discussion has a rather sluggish response. At present, the same sys- 
tem philosophy is being adapted to a hydraulic system. Because of 
the quicker response of the hydraulic system, more accurate results 
have been achieved, in that both velocity and acceleration errors have 
been reduced by a considerable amount. 

A stabilized mount for aerial cameras, which would keep the camera 
axis more nearly vertical, would result in considerable saving in the 
cost of aerial mapping projects. The gains from a good stabilization 
system would be twofold. First there would be the gain resulting 
from elimination of many cases where missions have to be reflown be- 
cause of extensive tilt in the photographs, and second, a gain in resolu- 
tion or sharpness of the picture by elimination of angular motion of 
the camera during exposure. 

The ultimate solution to the camera-stabilization problem can be 
broken down into two distinct parts. First would be the develop- 
ment of a system which would automatically hold the camera in a 
fixed attitude at all times. The second would be a system for deter- 
mining a true vertical reference axis and a follow-up system capable of 
trimming the camera axis always to be parallel to the vertical refer- 
ence. For practical usage, much is to be gained from a system which 
will retain the camera axis at a given tilt over a long period of time. 
Blur of the photographs resulting from roll or pitch of the airplane 
would be eliminated. Also, considerable saving in time and cost can 
be made in the rectification of the photographs. 

Based on experiments made in the stabilization of aerial gun tur- 
rets, we expect to produce a system of camera stabilization that will 
automatically maintain the camera's axis in a fixed attitude regard- 
less of tilts, rolls, pitches, or any maneuver of the aircraft. The sys- 
tem will have manual controls to permit trimming of the stabilized 
camera. Thus a manual adjustment of the camera's attitude can be 
made so that the stabilization system could be started at any time. 

92 LOCKNER Vol 49, No. 1 

Regardless of the original camera's attitude, the camera axis could be 
changed at will to make it as nearly vertical as possible within the 
limits of existing conditions. Precision spirit levels could be mounted 
on the camera to indicate tip and tilt so that by selecting a time when 
the aircraft was flying straight and level, a trimming operation could 
be made. Depending on the sensitivity of the system, any drift after 
erection would be appreciably small and result in a major gain for 
practical usage. 

We have received inquiries from time to time from professional 
people in the motion picture industry regarding their problems in re- 
gard to a stabilized camera mount. We feel that problems existent 
in that field very closely parallel those of aerial photography in many 
instances. It is safe to assume that the techniques of the stabilization 
system described above could very well be applied to the stabilizing of 
studio camera mounts. It is probable that a practical system from 
the point of view of cost, weight, space, and performance could be 
developed successfully. 



PATENT ATTORNEY: Warner Bros. Studios, Burbank, Calif., 
has opening for young patent attorney familiar with modern techniques 
in motion picture production equipment, color photography, sound re- 
cording, radio, and television. Give full details of background. Write 
N. Levinson,, Warner Bros. Studios, Burbank, Calif. 

PATENT DRAFTSMAN: Warner Bros. Studios, Burbank, Calif., 
desires draftsman skilled in electrical and mechanical drafting for patent 
purposes; knowledge of modern mot on picture equipment, including 
cameras, sound recording and reproducing, motion picture projection, 
radio, color, and television desirable. Give full details of background. 
Write N. Levinson, Warner Bros. Studios, Burbank, Calif. 


Carrying out the Society's current expansion program, which was started by 
Harry Smith, Jr., recently resigned Executive Secretary, the Society has made 
several changes in its Headquarters Staff and in the organization of its offices. 
Boyce Nemec, former Engineering Secretary, has been appointed Executive 
Secretary, and his office staff has been increased by the addition of a full-time 
Editor and a Staff Engineer. In addition, responsibility for the business and 
financial records and physical operation of the Society's Headquarters have 
been formally assigned to an Office Manager. 

Executive Secretary 

Boyce Nemec, newly appointed Executive Secretary of the SMPE, has been 
active in the technical end of motion pictures for over a decade. He was a member 
of the Visual Education Department of the University of Minnesota before joining 
the Army in 1941, when he was assigned to the Signal Corps Training Film 
Production Laboratories at Fort Monmouth, New Jersey. 

Mr. Nemec was instrumental in forming the War Committee on Photography; 
served as secretary of the Interim Armed Forces Committee on Photography in 
its initial stages; and represented the Signal Corps' engineering and procurement 
interests on the War Committee, Federal Specifications Committee and Joint 
Army-Navy Specifications Board as^chief of the Signal Corps' Photographic Speci- 
fications Unit. 

On his discharge from the Army, Mr. Nemec came directly to the SMPE as 
Engineering Secretary to carry out the Society's greatly enlarged standardization 
and engineering program. 


Office Manager 

Margaret C. Kelly was born in Wilkes-Barre, 
Pennsylvania, and attended Wyoming Seminary 
and Bucknell University. She came to the 
SMPE over four years ago to be Financial 
Assistant. In line with the Society's expansion 
program, Miss Kelly recently was made Office 
Manager. Before joining the staff of the SMPE 
she was Office Manager for .the Lion Chemical 

Staff Engineer 

Thomas F. Lo Giu- 
dice, Staff Engineer, 
received his motion 
picture experience in 
the engineering de- 
partment of the Inter- 
national Projector 
Corporation . There he 
did experimental and 
development work on 
motion picture projec- 
tion and sound systems 
and on naval under- 
water sound equip- 
ment. Later he served 

in the electronics 
branch of the U. S. 
Coast Guard dealing 
with radio, radar, so- 
nar, and loran gear. 
He recently received 
the B.E.E. degree from 
the Polytechnic Insti- 
tute of Brooklyn, 
where he was chairman 
of the AIEE branch. 
Mr. Lo Giudice is a 
member of the SMPE, 
IRE, and Eta Kappa 


Helen M. Stote was born in Colorado Springs, 
Colorado. She received the B.A. degree from 
Stanford University and the M.A. degree from 
the University of Wisconsin. Before coming to 
the SMPE as Editor, Miss Stote was Publica- 
tions Manager of the Institute of Radio Engi- 


Vol 49 AUGUST 1947 No. 2 


Photometric Calibration of Lenses Preface 


Compensation of the Aperture Ratio Markings of a 
Photographic Lens for Absorption, Reflection, and 
Vignetting Losses IRVINE C. GARDNER 96 

An Instrument for Photometric Calibration of Lens 
Iris Scales M. G. TOWNSLEY 111 

A Simplified Method for Precision Calibration of 
Effective / Stops F. G. BACK 122 

Remote Control and Automatic Focusing of Lenses 

H. C. SILENT 130 

Some Engineering Aspects of Amateur Projection 
Equipment for the Mass Market PERCIVAL H. CASE 139 

Proposed Standard Specifications for Flutter or Wow as 
Related to Sound Records 


Proposed Standard for 35-Mm Flutter Test Films 


Catalog of Research Council and SMPE Test Films 162 

Recent American Standards on Motion Pictures 171 

62nd Semiannual Convention 181 

Current Literature 185 

_>yrighted, 1947, by the Society of Motion Picture Engineers, Inc. Permission to republish 
iterial from the JOURNAL, must be obtained in writing from the General Offide of the Society. 
Society is not responsible for statements of authors or contributors. 

Indexes to the semiannual volumes of the JOURNAL are published in the June and December 
issues. The contents are also indexed in the Industrial Arts Index available in public libraries. 











** President: LOREN L. RYDER, 

5451 Marathon St., Hollywood 38. 
**Past-P resident: DONALD E. HYNDMAN, 

342 Madison Ave., New York 17. 
** Executive Vice-President: EARL I. SPONABLE, 

460 West 54th St., New York 19. 
^Engineering Vice-President: JOHN A. MAURER, 

37-01 31st St., Long Island City 1, N. Y. 
**Editorial Vice-President: CLYDE R. KEITH, 

233 Broadway, New York 7. 
* Financial Vice-President: M. RICHARD BOYER, 

E. I. du Pont de Nemours & Co., Parlin, N. J. 
** ''Convention Vice-President: WILLIAM C. KUNZMANN, 

Box 6087, Cleveland 1, Ohio. 
** Secretary: G. T. LORANCE, 

63 Bedford Rd., Pleasantville, N. Y. 
^Treasurer: E. A. BERTRAM, 

850 Tenth Ave., New York 19. 

**JOHN W. BOYLE, 1207 N. Mansfield Ave., Hollywood 38. 

*FRANK E. CARLSON, Nela Park, Cleveland 12, Ohio. 

*ALAN W. COOK, Binghamton, N. Y. 
**ROBERT M. CORBIN, 343 State St., Rochester 4, N. Y. 
**CHARLES R. DAILY, 5451 Marathon St., Hollywood 38. 
"tjAMES FRANK, JR., 356 West 44th St., New York 18. 

*JOHN G. FRAYNE, 6601 Romaine St., Hollywood 38. 
**DAVID B. JOY, 30 East 42d St., New York 17. 

*PAUL J. LARSEN, 1401 Sheridan St., Washington 11, D. C. 

*WESLEY C. MILLER, MGM, Culver City, Calif. 
**HOLLIS W. MOYSE, 6656 Santa Monica Blvd., Hollywood. 
*tA. SHAPIRO, 2835 N. Western Ave., Chicago 18, 111. 
*WALLACE V. WOLFE, 1016 N. Sycamore St., Hollywood. 

*Term expires December 31, 1947. tChairman, Atlantic Coast Section. 
**Term expires December 31, 1948. JChairman, Midwest Section. 
"Chairman, Pacific Coast Section. 

Subscription to nonmembers, $10.00 per annum; to members, $6.25 per annum, included in 

their annual membership dues; single copies, $1.25. Order from the Society at address above. 

A discount of ten per cent is allowed to accredited agencies on orders for subscriptions 

and single copies. 
Published monthly at Easton, Pa., by the Society of Motion Picture Engineers, Inc. 

Publication Office, 20th & Northampton Sts., Easton, Pa. 

General and Editorial Office, Hotel Pennsylvania, New York 1, N. Y. 

Entered as second-class matter January 15, 1930, at the Post Office at Easton, Pa. 

under the Act of March 3, 1879. 


Vol 49 AUGUST 1947 No. 2 



During the past few years there has been a rapidly growing need for 
a more accurate expression of the photographic speed of a lens than 
is afforded by the simple /-number ratio. 

The density of a photographic image depends on (a) the brightness 
of the subject, (b) the effective speed of the lens, (c) the speed of the 
film, and (d) the exposure time. In modern motion picture photog- 
raphy, all these factors except (b) are controlled and known to within 
a few per cent, but the supposed speed of the lens may be in error by 
as much as 60 or 70 per cent. This is caused by light lost by surface 
reflections or by direct absorption in the lens, and to incorrect marking 
of the /-number scale. 

For this reason the SMPE Standards Committee has recently set 
up a Subcommittee on Lens Calibration to study the whole subject 
and to recommend a standard procedure for measuring the effective 
photographic speed of a lens. The Subcommittee will also attempt 
to standardize a new system of speed markings which eventually will 
replace entirely the /-number markings that have heretofore been 
the only indication of lens speed. 

The urgency of this problem is borne out by the fact that since 1941 
no less than five papers have appeared in the SMPE Journal dealing 
with lens calibration, and three more similar papers are being pub- 
lished in the present issue. The new Subcommittee contains seven 
members from the West Coast, to represent the Hollywood interest, 
and seven members from manufacturing firms in the East and Middle 
West and from the National Bureau of Standards. 

* Chairman, Lens Calibration Subcommittee. 






Summary. At present the diaphragm markings of a photographic lens are based 
entirely upon geometrical considerations and do not take into account the losses of 
light resulting from absorption, reflection, and scattering. A method of equivalent 
marking is described in which, for example, the marking 8 does not correspond to the 
geometrically determined aperture ratio 1:8 but to an opening sufficiently larger to 
permit the transmission of as much light as would be transmitted by the aperture 1:8 
in the absence of any losses due to absorption, reflection, and vignetting. Such a sys- 
tem of apertures may be referred to as equivalent or compensated apertures. 

Two systems of compensation are given, one based upon the illumination at the 
center of the field and the other based upon the average illumination over the entire 
field thus taking vignetting into account. A relatively simple photometric procedure 
for determining either of the two systems of compensated graduation is described. 
For use during a transition period a system of markings is described which will permit 
exposures to be determined either with light losses compensated or by the present 
method without compensation. Except for the change of markings on a lens no other 
instrumental changes are required to apply the new system of exposure determination. 


Photography is both an art and a technique. Until a few years 
ago the technique was largely empirical, each photographer's practice 
being based on his own experience with his own particular equip- 
ment. Now, however, the principles underlying photographic 
technique are so well known and quantitative relations are so pre- 
cisely established that photographic engineering may be said to have 
become an applied science. The possibility of saving large sums of 
money and at the same time securing greater uniformity of results 
by greater precision in the exposing and processing of the very large 
quantity of film used for motion picture photography has constituted 

* Presented, Apr. 24, 1947, at the SMPE Convention in Chicago. 
** National Bureau of Standards, Washington, D. C. 


an economic urge in favor of this transformation. The development 
of the modern photoelectric exposure meter and the more general 
dissemination of quantitative information regarding the properties 
of emulsions have made it possible not only for the professional, but 
also for the skilled amateur to control his photographic work in a 
scientific manner. 

With this progress a demand has arisen for a more scientific method 
of marking the diaphragm openings of a lens. The method now in 
general use is based entirely upon the diameter of the entrance pupil 
and the equivalent focal length. It gives no consideration to the 
loss of effective light which arises because of absorption in the glasses 
of which the lens elements are made or because of reflection and 
scattering at the surfaces. Photographic objectives in common use 
may have from 4 to 10 glass-air surfaces. The transmissions of such 
lenses may range from 50 to 80 per cent if uncoated and may be as 
high as 95 per cent if efficient low-reflection coatings are applied to 
the surfaces. It is evident, therefore, that the effective exposures 
for different lenses with the iris diaphragm set for the same aperture 
ratio may vary by a factor of almost 2, an uncertainty which is in- 
consistent with the precision with which the other factors governing 
exposure are controlled. 

Several methods for calibrating and marking the apertures of a 
photographic lens which will effect a correction for the varying light 
losses with different lenses have been proposed. A method is pre- 
sented which is relatively simple and direct, and by which different 
laboratories may be expected to arrive at the same system of marking 
and equivalent values without the interchange of physical standards. 
The method is extended to apply to lenses when focused for infinity 
or for any finite distance, and a system for marking the diaphragm 
scale is suggested. 

In order to distinguish between the older and newer systems of 
marking, the aperture ratio of the older system, based on geometrical 
measurements only, will be referred to as the geometrical aperture 
ratio. The new aperture ratio, which takes the absorption and 
reflection losses into consideration will, for present purposes, be 
designated the equivalent aperture ratio. Other terms which might 
be used are compensated aperture ratio, and t aperture ratio, the i 
standing for transmission as suggested by Berlant. 1 " 3 Consideration 
is also given to the difference in effective exposure of different lenses 
because of vignetting. 

98 GARDNER Vol 49, No. 2 


In Fig. 1, L represents a photographic lens which receives light 
from the uniformly illuminated screen CD. In the focal plane there 
is the plate GH, with the aperture of area A centered at 0. Behind 
this aperture there is a photoelectric cell with receiver R which is 
large enough to receive all the luminous flux from the screen CD that 
is transmitted by the photographic lens and the aperture at O. The 
full lines represent the boundary of the cone of rays proceeding .from 

FIG. 1. Arrangement of apparatus for measurement 
of total luminous flux transmitted through a photo- 
graphic objective and through an aperture of area A 
in the focal plane at O. The geometrical aperture ratio 
is l:l/(-2sin a). 

the screen to the axial point of the focal plane. In the object space 
the cone degenerates into a cylinder of which the diameter is d. In 
the image space the half angle of the cone is a, where a is defined by 
the equation* 

sin a = d/2/. (1) 

The aperture ratio of the lens for an infinitely distant object is 
1 if/d, where f/d is the/ number. It is evident that 


2 sin 


In Fig. 2, CD represents a uniformly illuminated screen identical 
with that of Fig. 1. The screen GH with aperture, of area A at 

* There is a temptation to write this equation tan a = d/2/, but if the lens is 
suitably corrected for coma, a necessary condition for a photographic objective, 
Eq (1) is correct. 

Aug. 1947 



and photoelectric cell with receiver R, are identical with the similar 
parts of Fig. 1. At //, at a distance e from the plane GH, there is a 
circular diaphragm of diameter such that the angle a in Fig. 2 equals 
the corresponding angle of Fig. 1. For the moment it will be as- 
sumed that the length e is identical with the equivalent focal length 
of the photographic objective of Fig. 1. If B is the brightness* of 
the screen CD and if one assumes a very small area A^4 at 0, normal 
to the axis of the lens and including the axial point, the- total luminous 
flux AF received by this area is given by the equation 

AF = B AA sin 2 a, (5) 

this equation being exact in the limit as A^4 approaches zero. If it 

FIG. 2. Arrangement of apparatus for measurement of 
total luminous flux transmitted through the aperture // 
and through the aperture of area A at 0. This flux is 
equivalent to that transmitted by a lens free from absorp- 
tion and reflection losses and with the aperture ratio 

* Throughout this discussion the terms illumination and brightness are em- 
ployed. These terms usually refer to measures of radiant energy evaluated in 
terms of the luminosity curve and apply strictly only when the energy receptor is 
the eye or a detector having the same sensitivity curve. Strictly speaking, when 
discussing photographic applications, one should use terms referring to radiant 
energy evaluated in terms of the wavelength sensitivity of the photographic emul- 
sion. This paper is concerned chiefly with ratios of two similar measurements of 
radiant energy and, inasmuch as optical glass, the only absorbing material con- 
cerned, shows little selectivity over the range of wavelengths under consideration, 
all equations are true, to a satisfactory approximation, whether they are consid- 
ered to apply to light (in the restricted sense) or to radiant energy affected by the 
photographic sensitivity curve. The justification for the use of the terms apply- 
ing to light is the greater because there has been no general agreement on terms 
to be applied to radiant energy evaluated in terms of photographic sensitivity. 

1 00 GARDNER Vol 49, No. 2 

is assumed that the actual area at for which the incident luminous 
flux is measured has a maximum radius r, such that the ratio r/e does 
not exceed 0.02, for most photographic purposes the equation 

F = BA sin 2 a (4) 

is sufficiently exact, where F is the total luminous flux incident upon 
the area A in the focal plane. For the arrangement of Fig. 2 the 
geometrical aperture ratio and the equivalent aperture ratio are identical 
because for the opening // there are no reflection or absorption losses. 
Referring now to Fig. 1, the total luminous flux F, falling upon 
the area A , will be given with sufficient exactness by the equation 

F' = BAk sin 2 , (5) 

k being the transmittance of the lens. 

The values of F and F' will be indicated by the photocell if it is 
suitably calibrated. It is at once apparent that if B, A, and a are 
held constant for the two measurements, the transmittance is given 
by the equation 

k = j. (6) 

On the other hand, suppose that the iris of the photographic lens 
or the diaphragm // of Fig. 2 is adjusted until F F'. Let a be 
the value of the half angle of the cone in Fig. 2 for which this equality 
is obtained. The aperture ratio corresponding to a is 1: (1/2 sin a\) 
and the/ number is I/ (2 sin a). For this adjustment it follows that 
the light transmitted by the photographic lens is equal to that; which 
would be transmitted by a lens of zero absorption and with the aper- 
ture ratio 1: (1/2 sin a). This method therefore, provides a means 
for calibrating a lens in such a manner that the absorption losses are 

Table 1 gives the values of a corresponding to values of the aper- 
ture ratios that are commonly represented on photographic shutters. 
-To calibrate a lens, therefore, it is necessary only to have a series 
of diaphragms which, for the length e, will correspond to the different 
required values of a. A photocell reading is taken with the arrange- 
ment of Fig. 2, after which the photographic lens is substituted for 
the diaphragm and the iris adjusted until the same reading is ob- 
tained. This setting corresponds to th.e given geometrical aperture 
ratio and zero absorption. If, on the other hand, one wishes to 
determine the equivalent aperture ratio corresponding to the maxi- 
mum indicated geometrical aperture ratio, it is necessary to secure 

Aug. 1947 



the balance between measurements of Figs. 1 and 2 by adjusting the 
diameter of aperture // in Fig. 2. 


Half Angles a Corresponding to Standard Aperture Ratios 

Aperture Ratio 

a Degrees 

Aperture Ratio 

a Degrees 











1:22. 6 3 






1:5. 6 6 


1:45. 2 5 






Some of the details and necessary precautions can now be profit- 
ably considered. It is not necessary that the length e of Fig. 2 be 
exactly equal to the focal length of the lens being tested provided 
that the correct ratio is maintained between e and the diameter of 
the aperture. It is desirable that e be approximately equal (within 
5 per cent is satisfactory) in order that the approximation introduced 
by the finite area A' shall not differ greatly for the two measurements; 
The area A must be definitely limited and, when the lens is used, it 
should be located accurately in the focal plane. A suitable way to 
achieve this is by the use of a metal plate in the focal plane with a 
circular aperture of the desired area and a photocell back of the 
aperture with a receiver large enough to receive all the light that 
passes through the opening. Such a plate is indicated GH in Figs. 1 
and 2. The dotted lines in Figs. 1 and 2 indicate the portion of the 
screen CD that contributes to the illumination at 0. It is evident 
that the contributing areas for the lens and for the aperture will be 
smaller and more nearly identical if the screen CD be brought close 
to the lens or diaphragm, respectively. 

It is, therefore, recommended that the screen CD be placed im- 
mediately in front of the lens and illuminated by transmitted light. 
Uniform brightness of the smaller area thus utilized is more easily 
obtained, and deficiencies in uniformity of illumination become less 
important as the areas utilized during the two measurements become 
more nearly identical. It is, of course, essential that the brightness 
of the screen have the same value for the measurement with the 
diaphragm and with the lens. It is also assumed that all parts of the 
screen CD contributing to either measurement obey Lambert's law. 

102 GARDNER Vol 49, No. 2 

In other words, a collimated beam or a surface giving specular re- 
flection should not be used. 

There are variations of the experiment that suggest themselves. 
For the evenly illuminated screen CD, an integrating sphere may be 
substituted. Furthermore, the directions of the light may be re- 
versed with the source at and the receiver at CD. In this latter 
case, if CD is not replaced by an integrating sphere, the receiver of 
the light-sensitive element must be large enough to receive all the 
luminous flux and it must be uniformly sensitive over the entire 
area. This requirement can be most easily met if the receiver is 
placed near the lens L or aperture // as the required size is then 
greatly reduced. 

Strictly speaking, the spectral quality of the light proceeding from 
the source CD when the measurements are made should be identical 
with that reflected from the object to be photographed, and the 
spectral sensitivity of the photocell should be identical with that 
of the emulsion to be used. When one considers the great variations 
usually present in the spectral quality of the light proceeding to a 
lens from an object to be photographed, it is evident that it is not 
practicable to fulfill this condition. Fortunately the absorption of a 
photographic lens is not particularly selective for different parts of 
the spectrum, and the values of k for a given lens will not differ 
greatly as the spectral distribution of the light illuminating the 
screen CD is changed. However, for precise work, standardization 
is desirable and it is accordingly suggested that tungsten lamps 
operating at a color temperature of 2360 K be used in conjunction 
with Wratten No. 79 filters. This givefe light having a color tem- 
perature approximately that of the noon sun (5400 K). The use of 
a controlled source facilitates intercomparison between measure- 
ments at different laboratories. Even in the absence of such a 
standardized source, if the screen CD is illuminated by tungsten 
lamps operating at normal voltage and if the photocell has a special 
sensitivity similar to that of the commercial exposure meter, the use 
of the equivalent aperture ratios, measured by the method of this 
presentation, will be much more precise and accurate than the use 
of the geometrical aperture ratios. 


According to the current method of determining aperture ratio, 
the diaphragm markings on the lenses will not yield consistent 


exposures from lens to lens. Either the speed ratings of the emulsions 
or the computing tables on the exposure meters, or both, may be 
considered as adjusted to give the correct exposure for some average 
value of ko typical of photographic objectives. For a lens having 
a value of ko smaller than this average value, the result will be 
underexposure and for a lens having a higher value (for example, a 
coated lens) the result will be overexposure. If this assumed value 
of ko were known, it would be possible to alter the speed rating of 
the emulsions to give correct exposure with lenses graduated to. read 
equivalent aperture ratios. To illustrate, if it were known that the 
value of ko is 0.76 for the average lens 4 is the basis on which the com- 
putation tables of a certain make of exposure meter rests, the 
speed rating of a photographic emulsion for use with the equivalent 
aperture ratios should be increased by the factor 1/0.76. Once this 
adjustment of speed ratings, has been made, exposures should be self- 
consistent for all lenses graduated in terms of the equivalent aper- 
ture ratio. 

FIG. 3. The aperture scale of a typical lens with the geo- 
metric aperture ratios indicated by the graduations and num- 
bers in the conventional manner. Dots (which preferably are 
in red) indicate the settings for the equivalent aperture ratios. 
To illustrate, the dot between the graduations 5.6 and 8 
corresponds to the equivalent aperture 1 : 8. 

Fig. 3 shows the diaphragm markings of a lens developed into a 
linear scale. Between each pair of graduations there is a dot. The 
indicated graduations with numbers correspond to the geometrical 
aperture ratios as now marked on photographic lenses. The dots 
correspond to the equivalent aperture ratios, each lot representing 
the equivalent aperture ratio of the same value as the next smaller 
geometric aperture ratio. To illustrate, the dot between 2.8 and 4 
corresponds to the equivalent aperture ratio 1:4 and similarly the 
dot between 5.6 and 8 corresponds to the equivalent ratio 1:8. Such 
dots in red have been used at times on lenses for the Leica camera.* 

This method of marking may not be entirely unambiguous as there 
is a possibility of allocating the dot to the incorrect one of the two 

* Paul C. Foote, of Bell and Howell Company, has mentioned this type of 
marking and has made photographs available of a Leica lens so graduated. 

104 GARDNER Vol 49, No. 2 

adjacent stop numbers. Consequently the marking shown in Fig. 
4 is suggested. In this case a line, preferably red, is drawn connect- 
ing the setting for the equivalent aperture ratio to the corresponding 
geometrical aperture ratio graduation. The length of this line 
indicates the extent to which a lens aperture must be opened beyond 
a given geometric aperture ratio to compensate for the loss of 
light resulting from reflection and absorption. Having a lens 
doubly marked in this manner is certainly an advantage during a 
period of transition when one is changing from the regular use of 
one set of markings to the other. In addition to marking the lenses, 
the only change required is the publication of a new set of speed 
ratings. During the transition period the manufacturers might 

FIG, 4. A preferred system of marking the equivalent 
aperture ratios in which a line (preferably red) connects the 
setting for a given equivalent aperture ratio with the geo- 
metrical aperture ratio having the same numerical value. 

well give two sets of speed ratings in their tables, one in black for use 
with the geometrical aperture ratios, and one in red for use with the 
equivalent aperture ratios. Even after the use of the equivalent 
aperture ratios has become general it might well be desirable to re- 
tain the double system of marking on photographic lenses because the 
geometrical aperture ratios apply more precisely to the depth of focus 
scales with which many cameras are now provided. 


The measurements of Section II have been concerned only with the 
illumination on the axis of the lens. *In Fig. 2, if the aperture O 
and receiver R are displaced the distance e tan /3 in a direction 
normal to the axis of the diaphragm, as shown in Fig. 5 one obtains 
a measure of the flux corresponding to an image point at an angular 
displacement from the axis. This will be less than the axial value 
because of the operation of the "cosine fourth-power law", which is 
a statement that the illumination in the field of a photographic 
lens varies as the fourth power of the cosine of the angular distance 
from the center of the field provided that the diameters of all ele- 
ments of the lens system are so great that the iris of the lens is the 
only part of the system that restricts the cone of transmitted rays. 

Aug. 1947 



Although this last restriction is seldom fulfilled over the entire field 
of a lens, it should be noted that, even when this condition is com- 
plied with, the cosine fourth-power law is an approximate rather 
than an exact statement. 

The ratio of the illumination at the off-axis position to the axial 
value gives a measure of the decrease of illumination as the image 
point moves from the center of the field outward. Let Fp and F 
be the fluxes measured, as shown in Fig. 2 and 5, respectively. Then 
the ratio Fp/F gives the ratio of. the exposures degrees from the 
axis and on the axis for an ideal lens with no absorption and no 



FIG. 5. Arrangement of apparatus for measuring the total 
flux transmitted through the aperture // and through the 
aperture of area A displaced from the axial position. The 
measurements made with arrangements indicated in Figs. 2 
and 5 give a measurement of the decrease of illumination for 
points off the axis when there is no vignetting. 

The ratio F$/F will approximately equal cos 4 /3. 
Similarly, for the arrangement of Fig. 1, the aperture and the 
receiver R may be displaced through the distance / tan and a 

* The term "vignetting" is ambiguous unless defined. It may reasonably be 
applied to include all the decrease of illumination that arises at an off-axis point 
in the image plane of a photographic objective, or the term may be used to apply 
only to the decrease of illumination which arises because of restrictive action of 
parts of the lens mount or lens elements and which is in excess of that necessarily 
occurring with an ideal lens. Following the custom of several writers, the second 
application is used in this discussion. In accordance with this interpretation, 
there is no vignetting for the system of Fig. 5. 

106 GARDNER Vol 49, No. 2 

measurement of the illumination made. It will be assumed that the 
iris is set for an equivalent aperture ratio the same as that for the meas- 
urements of Figs. 2 and 5. We now have four readings, F' , the 
axial value for the arrangement of Fig. 1 ; F' $, the reading for the 
arrangement of Fig. 1 with the measurement made /3 degrees from 
the axis; and F and F ft . Since the equivalent aperture ratios 
are assumed to be the same in all cases, F' = F . The ratio F'p/Fp 
gives a measure of the vignetting, i. e., the falling off in illumination 
beyond that attributable to an ideal lens. * However, the ratio F'p/ 
F' is the more useful ratio to the lens user because it gives directly 
the ratio of the exposure obtained at a point J3 degrees from the axis 
to that obtained on the axis. 

Different types of lenses differ greatly in the amount of vignetting. 
A lens system which is long in comparison with its focal length 
requires much larger elements than a short lens system if vignetting 
is to be avoided or reduced to a satisfactory value. Sometimes a 
lens system is such that the aberrations for the marginal parts of the 
field are excessive. Such a fault is rendered less apparent when 
there is considerable vignetting because the lens is, in effect, stopped 
down at the edge of the field much more than at the center. Exces- 
sive vignetting is sometimes found in folding hand cameras because 
the manufacturer, in order to make the camera compact, makes the 
lens elements small. If, for example, the lens is rated as //2, this 
relative aperture may apply only at the center of the field, the ex- 
posure falling off very sharply for the corners of the picture. For 
such a lens, the vignetting rapidly becomes less as the lens is stopped 
down and such an arrangement do3S not necessarily represent an 
undesirable compromise. The user has a compact camera without 
excessive vignetting for the aperture ratios that are usually used 
with modern rapid film and at the same time has high speed, at 
least for the central part of the picture, for the occasions when it is 
required. However, when exposures are made on color film with the 

* This is not strictly true because the illumination resulting from the aperture 
of Fig. 2 does not fall off exactly as the fourth power of the cosine. The dia- 
phragm, however, represents a convenient standard that can be reproduced 
without difficulty at different laboratories and it follows the cosine fourth-power 
law as closely as do most lenses in common use. The approximation is better 
for the smaller aperture and for points near the axis. To illustrate, for points 
distant 40 deg from the axis the departures are 7.6, 1.8, and 0.4 per cent for the 
aperture ratios 1:2, 1:4, and 1:8, respectively. 

Aug. 1947 



maximum aperture, such vignetting, because of the generally reduced 
latitude of color film, may be sufficient to detract from the effec- 
tiveness of the picture. 


The method of stop calibration of Section II compensates for the 
different transmittances of different photographic objectives and 
ensures equivalent exposures at the center of the image field for 
different lenses when used at the same equivalent aperture ratios. 
However, it does not distinguish between the behaviors of different 

FIG. 6. The aperture in the plate GH includesTthe 
entire field of view utilized when the lens is mounted 
for use in a Camera. The measurement of luminous 
flux obtained under this condition is characteristic of 
the average illumination over the entire field instead 
of the illumination at the center of the field as obtained 
by the arrangement of Fig. 1. 

lenses which arise because of the differences in vignetting. In Fig. 
6 the arrangement is the same as for Fig. 1 except that instead of 
measuring the illumination of a small area near the axis one measures 
the total flux received by an area in the focal plane identical with 
the picture frame. Similarly for the comparative measurement with 
an aperture only, the arrangement of Fig. 2 is modified to give a 
measurement of the flux received by the entire field. The equivalent 
aperture ratios, as before, are considered equal when the two flux 
measurements are equal. If e of Fig. 2 and / of Fig. 6 are not equal, 

108 GARDNER, Voi 49, No. j 

the angular subtenses of the two field stops at the center of the dia 
phragm and exit pupil of lenses must be equal when the two measure 
ments of flux are made. For photographic purposes the agreemenl 
will be satisfactory provided that the two field apertures are identica 
and the values of e and / do not differ by more than one per cent 

For a motion picture camera or a miniature camera the field aper 
ture required is small and it is generally not too difficult to make 
the measure as outlined. If the picture frame is large it may b< 
desirable to reverse the direction of travel of the light, using ar 
integrating sphere to illuminate the picture frame. The total flu> 
is then measured by a photoelectric element with receiver large 
enough to receive all the light that comes through the lens. B) 
this reversal of the direction of travel of the light, a photocell with i 
receiver smaller than otherwise can be used. 

If the iris of a lens is calibrated by the two methods, measuring 
central illumination only and measuring the flux received by the 
entire field, it will be found that the calibrations, in general, are 
different. This is understandable because the two methods are 
based upon different assumptions. In the method first described, 
a given equivalent aperture ratio corresponds to a definite illumina- 
tion at the center of the image field. According to the second 
method the use of the equivalent aperture ratio corresponds to a 
given average illumination over the entire field. For a lens which 
has a large amount of vignetting a given geometric aperture ratic 
will correspond to a smaller equivalent aperture ratio than for a lem 
with less vignetting. Both methods of calibration have advan- 
tages, and it is probable that the different standardizing groups in- 
terested in photographic procedure should consider carefully the twc 
methods and make recommendations governing their use. 


In all the foregoing discussion it has been tacitly assumed that 
the object to be photographed is at an "infinite" distance and that 
the image consequently will lie in the focal plane of the lens. This 
is the basis on which the values of the geometric aperture ratio are 
engraved on the lens mounts and it is an entirely satisfactory 
procedure for a large amount of photographic work. If, for example, 
the object instead of being at an "infinite" distance is only ten focal 
lengths away, the distance from lens to focal plane is only increased 
by 10 per cent of the equivalent focal, length, and for many 


applications the error in exposure resulting from using the values 
corresponding to the image in the focal plane will not be excessive. 

Lenses for copying purposes and some other types of lenses are 
habitually used with the object distant only a few focal lengths 
from the lens and in such instances it is highly desirable that the 
aperture ratios be marked for one or more selected object distances 
approximating those which actually will be employed in practice. 
The method of stop calibration can be readily extended to apply to 
this problem. 

Suppose, for example, that the selected object distance is 2/, cor- 
responding to the use of the lens for one-to-one copying. Referring 
to Fig. 1, the plate GH bearing the aperture will be moved back 
from the lens to the image plane corresponding to one-to-one copying. 
Referring to Fig. 2, the plane GH will be separated from the dia- 
phragm by approximately the same distance as for the lens. For 
this particular case, the separation would be twice the equivalent 
focal length of the lens. With this new spacing the diameter of 
the aperture IJ should be determined so that sin a has a value 
corresponding to a selected aperture ratio as given in Table 1 . Meas- 
urements are now made as before, the diaphragm setting of the lens 
being altered until the two flux readings are the same. When equality 
is obtained, the equivalent aperture ratio of the lens, for the one-to-one 
ratio, is equal to the geometrical aperture ratio of the diaphragm. 

If the iris markings are calibrated in this manner the lens will give 
the same exposure for a given equivalent aperture ratio and one-to- 
one copying as does the same lens or any other lens with the cali- 
bration for infinite distances when used on a distant object. For a 
finite distance the question again arises as to whether the calibration 
should be similar to that of Section II, with exposure at the center of 
the field as the criterion, or the method of Section IV in which the 
criterion is average exposure over the entire field. 


In the foregoing text, reference has been made to several papers 
dealing with this subject which' have been published. The method 
of calibration and marking herein proposed offers advantages not 
possessed by any one of the previously suggested methods, as follows : 

1. The standard aperture to which reference is made is an aper- 
ture of known diameter in a metal plate and therefore can be 
readily and independently produced by any laboratory. This 


facilitates the maintenance of consistent systems of graduation by 
different laboratories. 

2. Each calibration is essentially a substitution method in which 
the two values of flux to be measured are of approximately the same 
value. This largely eliminates errors arising from the nonlinearity 
of response of the photometric apparatus and eliminates the need for 
carefully calibrated filters. 

3. No condenser or collimator system is used. Hence the method 
does not involve the assumption that the distribution of energy in a 
collimated beam is uniform. 

4. A method of application is proposed which requires no modi- 
fication of present models of exposure meters. New film-speed 
tables are required but presumably the data already in the possession 
of manufacturers of exposure meters will be sufficient for the prepa- 
ration of these tables. 

5. A system of lens marking is proposed which permits exposure 
to be determined either by the conventional or new method. 

6. An extension of the method of calibration has been given which 
permits lenses to be calibrated for object distances other than infinity. 

7. Calibrated value may be based on brightness of image at the 
center of the field or average brightness of image over all the field, 
as may be considered preferable. 


1 BERLANT, E.: "A System of Lens Stop Calibration by Transmission", 
J. Soc. Mot. Pict. Eng., 46, 1 (Jan. 1946), pp. 17-25. 

2 DAILY, C. R.: "A Lens Calibrating System", /. Soc. Mot. Pict. Eng., 46, 
5 (May 1946), pp. 343-356. 

3 MURRAY, A. E.: "The Photometric Calibration of Lens Apertures", /. Soc. 
Mot. Pict. Eng., 47, 2 (Aug. 1946), pp. 142-151. 

4 GOODWIN, W. N., JR.: "The Photronic Photographic Exposure Meter", J. 
Soc. Mot. Pict. Eng., 20, 4 (Apr. 1933), pp. 95-118. 


CHAIRMAN LORANCE : I have a question that I would like to ask Mr. Gardner. 
Is there a real necessity for keeping the opening in your method at approximately 
the focal distance or is that a convenience? 

MR. I. C. GARDNER: If the screen that is furnishing your light were absolutely 
uniformly illuminated and large enough, there would be no need for it, but by 
doing that you are using approximately the same portion of the screen for both 

MR. F. G. BACK: I should like to say that in this method there is no problem 
of focus. After the meeting I would be glad to show any one who is interested 
how simple it is. 



Summary. An instrument for calibrating iris scales of photographic lenses is 
described. The iris scales are calibrated in T stops based on the photometric trans- 
mission parallel to the axis. This follows the procedure proposed by Daily. 1 The 
new instrument employs measuring and comparison of optical paths which receive al- 
ternate light pulses from a single incandescent source. Calibrated attenuation is 
provided in the comparison beam so that the two systems may be balanced to give a 
null output from the photomultiplier cell which is used as a null detector. The 
null-balance principle makes the unit extremely stable and the sensitivity is suffi- 
cient to make accurate measurements on iris openings as small as 0.031 in., which 
corresponds to 1 in. T 32. Data are given on transmission of several lens types, and 
on the accuracy of the instrument, and a proposal is made for changing over to the 
new system. 

There is growing interest in the photometric calibration of photo- 
graphic lens iris scales. Studio (motion picture) photography and 
amateur color film are placing an increasing premium on accuracy of 
exposure. Reflection-reducing coatings have increased the variation 
from lens to lens in the exposure produced by any given /stop because 
the upper limit of lens transmission has been raised to approximately 
95 per cent from the old maximum of perhaps 85 per cent; so that it 
is now possible to have nearly a two-to-one ratio between the ex- 
posures made with two lenses having the same geometrical / stop. 
There seems no alternative to the eventual adoption of a photometric 
system of calibration. 

Several methods of calibrating lens iris scales on a photometric 
basis have been proposed from time to time. Of these, the proposal 
advanced by Daily 1 seems the most logical and has the advantage 
that it is reproducible in any laboratory without interchange of any 
master standards. This method involves the comparison of the light 
flux transmitted from a collimated beam entering a fixed circular 
opening with the light flux transmitted from the same beam by the 

* Presented Apr. 24, 1947, at the SMPE Convention in Chicago. 
** Bell and Howell Co., Chicago, 111. 


112 TOWNSLEY Vol 49, No. 2 

lens under test. The T stop corresponding to any iris opening is 
the quotient of the focal length of the lens divided by the diameter 
of the fixed circular opening having the same light transmission. The 
T-stop system proposed by Daily will then have the same significance 
as the present /-stop system except that the transmittance will be 
corrected for and will employ the same series of numbers to desig- 
nate openings. A lens of a given T stop and no transmission loss in 
the lens would have a physical iris (strictly, entrance pupil) open- 
ing equal to the comparison opening and, therefore, would have/ 
stop equal to the T stop. 

If the transmittance is k, we have the following relationship be- 
tween T and / stops : 

f = L/d 

T = L/D 

where L is the focal length 

d is the entrance pupil diameter 
D is the comparison stop diameter 

= k 

T = L/dVk T = f/Vk f = 

Hence, if k becomes equal to 1 (100 per cent transmittance), the T 
stop and the / stop will be the same. In general, k will be less than 
1.0 and the opening will be larger for a given T stop than for an / 
stop having the same number, and will transmit more light by a 
factor 1/k. 

Gardner 2 * proposes to select a lower value of k so that lenses cali- 
brated in the T system will correspond more closely to present lenses 
and will result in a minimum of change in present exposure meters and 
depth of field tables. This seems unnecessary. The exposure cor- 
rection may be easily taken into account by using a simple multiplying 
factor l/k to correct the film-speed value to be used. For the Weston 
exposure meter, the factor is implicitly given in a paper by Goodwin 3 
as 1/0.76, since the value 0.76 was introduced into the exposure- 
meter equation as a correction factor for lens losses. Future ex- 
posure meters could use k = I in the equation t = kT 2 /Bos. 

The correction for depth of field tables is of even less importance 
because of the uncertainty and lack of general agreement on the circle 

* AUTHOR'S NOTE: In the published version of Gardner's paper, this proposa 
of K < I is omitted. 


of confusion to be used. Existing tables may be used without serious 
difficulty, or new tables may be based on corrected values of the circle 
of confusion. In general, the tables will indicate a 'slightly too great 
depth of field if existing tables are used with T-stop lenses. 

Support for the choice of k equal to 1.0 is shown in Fig. 1. This 
figure shows a comparison of the marked / stops of a group of 10 
lenses with actual T stops. The circular points represent the un- 
coated lenses in the group, while the squares represent the 
coated lenses. It will be seen that the coated lenses fall very close 
to the line = 1.0, while the uncoated lenses in general fall well 


II 1 1 

1 1 



l n 

. 1 


; ; i ; 







oj .8 







h- c 






J_ -4" 





II 1 1 


( | 


2.3 2.8 4.0 5.6 8.O II Ife 


FIG. 1. Transmittance of a group of lenses. 

below this line. This means that there are already many lenses in 
existence, marked in / stops where the value of k is very nearly equal 
to 1.0, and that properly coated lenses can almost reach this value. 
It will also be seen that the value of k varies widely, making the 
choice of any particular value different from 1.0 very difficult. 

There seems to be no important reason for showing both the / and 
T scales on any single lens. Such a dual marking has been used on 
one German lens which has been examined by the author, and has 
been described by Gardner from data supplied by the author. The 
confusion incident to the use of dual markings would far outweigh any 
possible advantage. It is considered quite important that the 



Vol 49, No. 2 

markings be placed at the exact points in the true \/2 series, even 
though the figures shown on the lens are rounded, e. g., 11 should be 

As an illustration of the reproducibility of the basic system of 
lens calibration proposed by Daily, we may cite a recent experience 
with the system. An order was placed with an English optical firm 
for a quantity of lenses which were required to be photometrically 
calibrated. It was decided that the method proposed by Daily would 
be employed and that the lenses would be marked in T stops according 






_| I 



4.0 5.6 



FIG. 2. Test of sysem reproducibility. 

to Daily's definition. At the time, Daily's paper had not been pub- 
lished. The present author, from the published abstract of Daily's 
paper and notes taken during its reading, wrote a description of the 
method which was forwarded to the English manufacturer. From 
this description, he was able to construct the basic apparatus (using 
a different light-measuring system), and furnish the lenses calibrated 
in T stops. 

On receipt of the lenses, they were rechecked in this laboratory and 
found to be in good agreement with our own measurements. The 
actual comparison is shown in Fig. 2. The data are plotted as relative 
exposure versus marked T stop, where the relative exposure is the 
ratio of the actual exposure which would occur to the exposure 


computed from the marked T stop. It will be seen from the figure 
that only one stop in the entire group of 20 lenses falls outside the x /4- 
stop limit lines proposed by Berlant. 4 

It is not the primary purpose of the present paper to consider 
in detail the system to be chosen, although it has seemed worth while 
to present the personal opinions of the author and his associates as a 
part of the introduction. The opinions of others are already on the 
record. 1 " 7 Rather, it is proposed to describe an instrument which 
combines a null-balance photoelectric measuring method described 
by Carpenter the sensitivity of the photomultiplier cell, and the basic 

FIG. 3. Schematic optical layout. 

method of Daily, into an instrument which is stable, precise, and 
sufficiently sensitive to calibrate stop openings accurately as small as 
1 /32 in. in diameter. The optical system is shown schematically in 
Fig. 3. In accordance with Daily, the instrument incorporates a 
lamphouse A in which a 750-w projection-lamp filament is imaged in 
a small (Vs-in. diameter) opening B which forms the source for a large 
collimating lens C. An opening D in an integrating sphere E faces 
this lens at a convenient distance. A holder is provided over this 
opening D into which slides perforated with standard openings may 
be inserted. Provision is made for mounting the lens P to be cali- 
brated in front of the sphere opening so that all of the light leaving 
the lens is transmitted into the sphere. An electron-multiplier photo- 
cell F is placed in the sphere wall at 90 degrees to the window. So 
much of the system follows Daily. 



Vol 49, No. 2 

From the rear side of the lamp filament, a second light beam is 
carried by mirrors and condensers to form a second filament image in 
the same plane as the collimator source image and approximately 6 
in . away at G. The fan motor which cools the projection lamp has its 
shaft horizontal and midway between these two images. A chopper 
wheel H is mounted on this shaft to interrupt both beams, and the 
apertures are phased so that the two beams alternate. The second 
beam is used as the comparison beam in a manner analogous to that 
used by Carpenter in the Baird nonrecording densitometer. 8 A lens 
J images the source aperture on an opal glass K. This lens is pro- 
vided with a series of fixed circular stops increasing in \/2 ratio in 

FIG. 4. Completed instrument. 

diameter, corresponding to the iris stops from T 2 to T 32. This lens 
and the stops serve as a stepped attenuator. A second lens M images 
the opal glass on the photocell, projecting its beam by means of a 
mirror directly across the sphere. The second lens has a wedge- 
shaped slit diaphragm N sliding across a fixed narrow slit to give 
continuous attenuation of the light intensity. This slit is calibrated 
in terms of the equivalent focal lengths of the lenses to be tested. 
An iris Q in the illuminating portion of the beam serves for initial 
balancing, and for rebalancing when lamps are changed. 

The construction of the instrument in its present form is shown in 
Fig. 4. The lamphouse is seen at the left end of the optical bench 
support, with the collimator lens between the lamphouse and the 


sphere. At the rear of the lamphouse are the condenser lenses, and 
the mirrors which furnish the light beam for the comparison track are 
mounted on the bench on the side away from the operator, with the 
lens having the circular stops closest to the lamphouse and the lens 
having the wedge attenuator closest to the sphere. The amplifier 
is placed where it can be reached for occasional adjustment of the 
gain control at extremely high or low light levels. 

Fig. 5 shows the instrument with a lens in place for direct photo- 
metric calibration and marking of its diaphragm ring. The lens is 
mounted in a separate holder in front of the sphere opening, and a flat 
steel table and a small scriber point are used to mark the iris points at 

FIG. 5. Completed instrument showing lens in place for calibration. 

the various T stops. In use, the focal length is set off on the wedge 
scale, the T stop is set into the first lens in the comparison beam, the 
lens to be calibrated is stopped down to bring the null indicator to 
balance, and the scriber is used to mark the point on the ring. This 
point is later picked up in the engraving machine and the permanent 
engraving done from the scribed line. 

The double interrupted beams give alternate pulses of light into 
the sphere, and the calibration beam may be attenuated to match 
the magnitudes of the two pulses to give zero variation in light in- 
tensity on the photocell. 

The electrical portion of the instrument is shown schematically 
in Fig. 6. A Type 931 A photomultiplier cell is used as the sensitive 



Vol 49, No. 2 

element. This cell has the S4 response, which peaks in the blue. 
When this tube is used with an incandescent lamp, the net effective 
response peaks at 500 millimicrons, and gives a fairly good compro- 
mise. Ideally, as Daily points out, in order to take coating color 
into account, the response should correspond to panchromatic film. 
Two cells of different response and appropriate filtering can be made 
to give a good approximation to this response, but there is no red- 
sensitive electron-multiplier photocell commercially available; 900 
volts direct current are supplied to the voltage divider which furnishes 
dynode voltages to the cell. This voltage is filtered by the 10-henry 
choke and the two l-/if capacitors to reduce 60-cycle hum, and one 


FIG. 6. Schematic circuit of null detector. 

of the capacitors is placed within the shield around the photocell to 
filter out any possible line pickup. Output voltage is fed through a 
shielded cable to a single-stage amplifier as shown. The voltage 
supply shown for this amplifier is slightly unconventional because of 
modifications to suit a conveniently available transformer which 
was designed to furnish power to several tubes, and had to be equipped 
with a bleeder and voltage divider to give the voltages required for 
the present amplifier. The multiplier output is coupled through the 
O.Oljuf -capacitor, and gain control is provided by a 4-megohm potenti- 
ometer, the voltage to the grid being determined by the position of 
the arm on this control. A parallel- T feedback loop is used to 
tune the single stage to pass only the 1080-cycle frequency to which 


the chopper wheel is set. This tuning is kept quite sharp to eliminate 
stray pickup and 60-cycle disturbance so far as possible. 

In general the circuit is very similar to that used by Carpenter, 
with the exception of the use of a pentode (6SJ7) for the amplifier 
tube and the tuning to a different frequency. Instead of the rec- 
tifier and tuning eye which Carpenter uses for a null detector, the 
present unit uses a 3-in. oscilloscope. There are two reasons for this 
choice. The first reason was the ready availability of the instrument 
in the laboratory. The second reason, and the one which really 
controlled the choice is that there is- a 120-cycle component in the 
light from the lamp which cannot be completely filtered out, and 
which, particularly at very high light levels, modulates the unbalance 
voltage. If the oscilloscope is swept at 60 cycles, by applying line 
frequency to the horizontal plates, this 120-cycle modulation appears 
as a stationary loop on the screen with the unbalance voltage super- 
imposed as modulation on the loop. This very considerably increases 
the precision with which a null setting can be made. 

If the instrument were to be restricted to lower light levels within 
the sphere by limiting its range or by attenuating the primary beam, 
it might be possible to omit this refinement and use the tuning eye 
as a null indicator. It was not considered desirable to employ at- 
tenuation in the primary beam because of the necessity of main- 
taining calibration of the attenuator, and it seemed essential to cover 
at least the range of diameters from 1 / 32 to 2 in.. When it is realized 
that this represents a l-to-30,000 range of light intensity within 
the sphere the difficulty of maintaining satisfactory sensitivity over 
this range will be appreciated. Attenuation of the primary beam by 
a nonselective means which would not change the optical character- 
istics of the system might be satisfactory for a production-type 
instrument. Evaporated inconel films as described by Benford 
might be satisfactory for this purpose, since such films have uniform 
absorption for all colors of the visible spectrum, and are nonscattering. 

Original calibration of the instrument is made by means of a set 
of fixed, accurately made circular stops ranging in size from 0.0312 to 
2 in. by steps having a ratio of 2 in area. These stops are placed in 
the holder over the sphere window, and the stops in the first lens in 
the comparison beam are matched to them, point for point, for at 
least two settings of the wedge diaphragm. The wedge diaphragm 
is then calibrated for equivalent focal length in the same manner, for 
at least four combinations of primary and secondary stops. Once 

120 TOWNSLEY Vol 49, No. 2 

the instrument has been calibrated in this way, it is necessary only 
to check one or two points occasionally to be sure that the calibration 
remains correct. The null-balance method eliminates the effects 
of electrical drift, and only movement of elements or dirt in the op- 
tical system can affect the calibration of the beam-attenuating system. 

Measurements are made by inserting the lens to be tested in the 
collimated beam in front of the sphere window at P. If the lens is 
being originally calibrated, the focal length is set off on the wedge 
aperture scale, and the desired T-stop value inserted in the compari- 
son beam in the first lens /, the iris of the lens being graduated is 
closed until the null point is reached, and the point is scribed on the 
iris ring. This procedure is repeated for each stop to be marked. 

If the lens under measurement is being checked for correctness of 
existing engraving, the corresponding stop is inserted in the com- 
parison beam, the iris is set to the mark, and a balance is secured by 
varying the wedge aperture. The corresponding equivalent focal 
length is read from the calibration curve for the wedge, and the T 
stop is computed from the relation 

indicated T X true L 

true T 

indicated L 

We have here the situation where a lens of focal length L has a 
marked T stop T\ and hence an equivalent diameter D\. The cali- 
bration beam is therefore set to balance the light transmission of a 
stop of diameter DI. But experimentally, we find that the stop 
marking is in error and shifts the focal-length slide to a new position 
corresponding to a focal length LM- This changes the equivalent 
opening diameter for which the system is balanced to 

DM = 

but T M = 

whence L/T M = L M /T t 

and TM- = TjL/L M . 

The system of T stops proposed by Daily, and followed in this work, 
requires an accurate knowlege of the equivalent focal length of the 
lens being calibrated. We are fortunate in having available to us, in 
our Optical Engineering Department, a focal-length collimator with 
which all focal length measurements used in this work were made. 
This collimator consists of a well-corrected lens of 24-in. focal length 
having a ruled reticle in its focal plane. Rulings on this reticle 
subtend accurately known angles. The lens to be measured is 


set up facing the collimator, a microscope is focused on the image 
plane, and the lateral spacing between images of the rulings is 
measured. If the angle subtended by two rulings is 6, and the meas- 
ured distance between their images is d, the equivalent focal length is 
found from the equation 

L- d 

2 tan 6/2' 

This method is identical in principle with the one shown by 
Daily, but is much more convenient and rapid if the collimator is 
available. The collimator method is given in detail by Hardy 
and Perrin 9 . 

Where large numbers of lenses are to be calibrated to commercial 
standards of accuracy, it is expected that the labor of measuring 
individual focal lengths can be avoided by using the group average 
focal length for the entire group providing the group does not show 
excessive scattering of the individual focal lengths about the average. 
It is usual experience for all lenses of a given, design to group quite 
closely in focal length, even though the group average may be some- 
what different from the nominal or design focal length. This is true 
even for batches of a given design manufactured at different times. 
The effect is probably caused by failure of a designer to readjust the 
design when the focal length happens to differ slightly from the 
nominal figure after all of the aberrations have been corrected, and 
by slight further changes caused by fitting the design to existing 
curves in setting up the tooling for the lens. Once the tooling is 
fixed, there is little likelihood of the focal length shifting during 
manufacture except as it is slightly affected by tolerances in index, 
thickness, and spacing. 

The permissible accuracy of the T stop governs the tolerance on 
the measurement of the equivalent focal length. Where production 
lenses run within a total tolerance of 2 per cent, the batch average 
may safely be used as a basis for marking the T stops. The batch 
average may be as much as 8 to 10 per cent above or below the nomi- 
nal focal length. This will in general introduce an error too great to 
be tolerated into the T-stop markings if the nominal focal length is 
used as the basis of calibration. 

This point is covered here in some detail to emphasize the necessity 
for using at least the group average focal length and not using the 
nominal or marked focal length on which to base the calibration. 

122 BACK Vol 49, No. 2 


1 DAILY, C. R.: "A Lens Calibrating System", /. Soc. Mot. Pict. Eng., 46, 5 
(May 1946), p. 343. 

2 GARDNER, IRVINE C.: "Compensation of the Aperture Ratio Markings of 
Photographic Lenses for Absorption, Reflection, and Vignetting Losses", (pre- 
sented, Optical Society of America Fall Meeting, 1947); /. Soc. Mot. Pict. Eng., 
49, 2 (Aug. 1947), p. 96. 

3 GOODWIN, W. N., JR.: "The Photronic Exposure Meter", /. Soc. Mot. Pict. 
Eng., 20, 2 (Feb. 1933), p. 95. 

4 BERLANT, E.: "A System of Lens Stop Calibrations by Transmission", 
/. Soc. Mot. Pict. Eng., 46, 1 (Jan. 1946), p. 17. 

5 BACK, F. G.: "A Simplified Method for Precision Calibration -of Effective / 
Stops", /. Soc. Mot. Pict. Eng., 49, 2 "(Aug, 1947), p. 122. 

6 MURRAY, A. E. : "The Photometric Calibration of Lens Apertures", /. Soc. 
Mot. Pict. Eng., 47, 2 (Aug. 1946), p. 142. 

7 SILVERTOOTH, E. W. : "Stop Calibration of Photographic Objectives", /. Soc. 
Mot. Pict. Eng., 39, 2 (Aug. 1942), p. 119. 

8 CARPENTER, R. O'B.: "New Light-Balancing Circuit for the Non-Recording 
Densitometer", J. Opt. Soc. Am., 36, 11 (Nov. 1946), p. 676. 

9 HARDY, A. C., AND PERRIN, F. H., "Principles of Optics", McGraw-Hill 
Book Co., New York, N. Y., 1932, Chap. 19. 


F. G. BACK ** 

Summary. Many methods have been proposed to replace the geometrical f ratio 
by an effective calibration which takes the transmittance- of the lens into consideration, 
but no method proposed so far has been accepted. The author outlines a new method 
of photoelectric lens calibration using the null method which compares the lens to be 
calibrated to a standard lens. This standard can be easily reproduced with a very 
high uniformity of performance. The method has been tested very thoroughly and is 
now in practical use for lens calibration. The accuracy obtained with this method 
is far superior to any procedure used so far. Its simplicity enables even an unskilled 
operator to obtain accurate results combined with a high working speed. 

It is a well-known fact that our present method of designating lens 
/ stops by the ratio of the focal length to the diameter of the entrance 
pupil is not satisfactory. In this age of multisurfaced lens systems, 
the geometrical / ratio is a very unreliable measure for the amount of 

* Presented Apr. 24, 1947, at the SMPE Convention in Chicago. 
** Research and Development Laboratory, 381 Fourth Ave., New York, N. Y. 


light reaching the film. It is not necessary to go into the reasons 
therefor; they have been explained time and again. It is certainly 
annoying if, relying on the speed calibration of our lens, we find out 
after processing, that the exposure is incorrect and that the 
lenses of different focal lengths give different exposures in spite of 
being set to the same / stops. The reason therefor lies in the fact 
that the geometrical /ratio tells us only about the amount of luminous 
flux entering a lens system while the exposure is determined by the 
amount of light passing the lens. To avoid an unpleasant surprise 
of the kind mentioned above, all lenses should be calibrated by 
transmission ; then we shall be assured of proper exposure, according 
to our meter, regardless of focal length and origin of our lenses. 

Many excellent methods have been proposed to measure the actual 
light transmission of photographic lenses but none of these methods 
has been accepted generally, because, while scientifically correct, they 
can be used only by highly skilled operators and require extensive 
and complicated apparatus. They can, therefore, be used only in 
lens-manufacturing plants or laboratories where the necessary equip- 
ment and personnel are available, but they are beyond the scope of 
an average motion picture studio, to say nothing of small production 
units, photographic dealers, and repair shops. 

Any new lens-calibration method should take into consideration 
the vast number of lenses presently in use and not only the lenses to 
be manufactured in the future. Such a method should, therefore, 
be so simple to operate that all these groups mentioned above can 
easily and speedily check and, if necessary, recalibrate the lenses 
they have on hand. 

A procedure is not generally usable if it requires the determination 
of the exact equivalent focal length for each individual lens, as a 
prerequisite for the calibration measurements. Also, methods re- 
quiring amplifiers, attenuators, chopper wheels, oscilloscopes, or 
even successive measurements are impracticable, because they entail 
the danger of objective and subjective measuring errors. To obtain 
a reliable result by such methods, a great number of measurements 
for each stop is required and the observational and operating errors 
have to be eliminated by curve-fitting, partial correlation, and other 
highly complicated mathematical methods which also are far beyond 
the reach of the average user of photographic lenses. 

In developing our method for light-transmission calibration, we 
were guided by the facts and requirements mentioned above, We 

124 BACK Vol 49, No. 2 

wanted to simulate as closely as possible actual camera conditions. 
Our main aim was uniformity of exposure, regardless of focal length 
or design of lens, easy and exact reproducibility of the measuring 
setup, and complete independence of the skill of the operator and 
measuring conditions. The proposed standard may seem arbitrary 
at first glance, and perhaps does not meet the requirements of a 
scientifically exact standard of comparison, but the new effective 
/ values do not deviate too much from the /-stop figures now in use. 
Any substantial deviation from these / values would not only render 
all meters, depth scales, and other tables based on the present / ratio 
obsolete, but it would cause a great deal of confusion among profes- 
sional and amateur photographers and cinematographers alike. 
Especially in the motion picture industry, the / stop does not only 
designate the aperture of the lens but is also used to measure the 
lighting of the set. Finally we think that the scope of a lens-cali- 
bration method should not be confined to motion picture lenses only, 
but should also include all photographic lenses. 

Summarizing the requirements of a lens calibration method as 
described above, we find the following : 

1. The accuracy obtainable should be better than 5 per cent. 

2. The instrument should be independent of observational 

3. It should be independent of current fluctuations or other 
operating conditions. 

4. The instrument should be simple, with no mechanical or 
electrical parts such as amplifiers or sensitive meters which can get 
out of order easily. 

5. The apparatus should be reproducible with commercially 
available material anywhere, anytime; and each piece of apparatus 
built to specifications should work within 5 per .cent limit. 

. 6. The device should be simple to operate so that even an 
unskilled person should be able to calibrate or measure with a high 
degree of accuracy. 

7. The method and the instrument should be generally usable 
for all photographic lenses and not only for 35-mm motion picture 

Other methods proposed so far have met some of the requirements 
but there is no method, to our knowledge, which meets all seven of 
the requirements listed above. 

Aug. 1947 



Requirements 1, 2, and 3 can be obtained by a null method only. 

Requirements 4 and 5 can be achieved only with the barrier- 
layer cell instruments which work without amplifiers. 

Requirement 6 can only be obtained by using a comparison 

Requirement 7 has not much to do with the method itself, but 
depends largely on the mechanical design. 

$o far, each of these principles by itself has been applied previously 
in other methods, but they have not been combined because there 
were certain links missing. 




FIG. 1. 

The instrument we designed to meet the abovementioned speci- 
fications is a very simple and compact one and is illustrated in Fig. 
1. The instrument consists of four basic parts: (1) light-source 
assembly; (2) variable standard assembly; (3) fixed standard and 
calibrating assembly; and (4) balancing meter assembly. 



Vol 49, No. 2 

The light source A which is a monoplane 500-w incandescent 
projection bulb throws two opposite light beams over two condenser 
systems on two opal-glass disks, BI and B 2 . The condensers are 
arranged in a way to give even illumination over both disks. These 
disks are always equally illuminated independent of voltage drop in 
the power-supply line and present a balanced light source for the two 
standards, C\ and G. 

Standard Ci is a variable standard which contains a disk with 
a number of exact calibrated aperture holes. Standard C 2 'is a 
fixed standard with an /stop of //4. Standard C 2 serves only as a 
balancing standard before the actual calibration is done and is re- 
placed by the lens to be calibrated. Both standards are mounted on 













. i, 






* '' 

* A 








Aperture Size 
1 in. 
3 A in. 
V in. 
1 A in. 

FIG. 2. Comparison Standards. 

movable carriers, A and D* t which allow for changing the distance* 
-between light source and lenses. Both carriers contain cells. The 
photoelectric cells are connected with a sensitive meter E which 
shows zero if both cells are in light balance. A potentiometer is 
used to compensate for the small deviations in the symmetrical 

We propose as a standard two identical plano-convex lenses made of 
BSC-2 glass, placed with their vertexes against each other with the 
aperture stop in the middle, as indicated in Fig. 2. 

We dislike going into mathematical details but those dimensions 
have the advantage of being inch fractions in so far as aperture holes 

Aug. 1947 



are concerned and can be reproduced with commercially available 
tools. All millimeter measurements given there can be measured 
with an ordinary micrometer, except for the radius of the lens cur- 
vature, which has to be done with a standard spherometer. 

Using this suggestion as a basic standard for lens-transmission 
calibration, a diaphragm hole of one in. represents a geometrical 
aperture of //2, an aperture diameter of J /2 in. represents a geometrical 
aperture of //4, J /4 in. represents //8. This standard lens has, of 
course, a certain transmission loss caused by reflection, aberration, 
and absorption. We do not care how big the percentage of trans- 
mission loss is but we shall take this transmission loss caused by the 
physical properties with this particular standard, as a base for our 
comparison method. This standard can be reproduced anywhere, 
at any time, with a high degree of accuracy because BSC-2 glass is very 
uniform. Refraction index and dispersion are everywhere held within 
very close limits. A well-polished surface has always the same re- 
llectivity. There is no coating applied which might cause variations. 
There are no cemented surfaces introduced which cause a different 
percentage of absorption, and last, but not least, this particular glass 
is readily available. This stand- 
ard has similar light 1 - transmission 
properties to a large number of 
commercially available lenses 
now in use, as for instance, the 
Eastman Kodalf //2.7 series. 
This presents a great advantage, 
since many lenses will not have 
to be recalibrated because their 
present calibration is very close 
to the /-stop marking of our 

As mentioned previously, it is 
necessary to use self-generating 
photocells which do not require 
an amplifier. Since two are 
needed for the comparison 
method, both cells should be 
equal as far as characteristics 
and sensitivity are concerned. 
Pairs of equal cells are easily FIG. 3. 







Vol 49, No. 2 

obtainable; but there is one disadvantage. Since they are equal 
only as long as the sensitive surfaces are illuminated evenly and 
over the entire cell, which is 40 mm in diameter (and smaller cells 
are not available), we had to find a method which distributes the 
flux of a smaller surface evenly over the entire cell surface. Here, a 
simple little invention was made consisting of a frustum of a cone, 
made of BSC-2 glass, with a base 40 mm in diameter, and a top 
surface of the area to be measured. This little integrator is silver- 


O 8 MM 


35 MM 



FIG. 4. 

plated on the outside, except for the base and top. The base is 
polished and the top is left fine-ground. (See Fig. 3.) Experiments 
have shown that a 25-deg cone gives optimum results. By intro- 
ducing two of these cones into the measuring instruments, illumina- 
tion of two small surfaces could be compared with a very high de- 
gree of accuracy, with commercially available pairs of photoelectric 
cells and with a standard microammeter. 

The next point to be considered is the size of the surface to be meas- 
ured. There are many differences of opinion on this subject; all of 
them have their merits and their disadvantages and advantages. 

Aug. 1947 



Some authors say we should measure only the center of the field; 
some like to measure the mean of the entire frame and some of them 
wish to measure just a fraction of the frame. In order to come to 
some conclusion and not to be too arbitrary, we made practical tests. 
Slides were made with different over-all density and different density 
distribution. Some of them had the same center density, some had 
the same over-all density, and some of them were unevenly illumi- 
nated. These slides were projected in different sequences on a screen 
and a number of observers gave their opinions on which slides rep- 
resented slides of equal density. Considerable density variations 

FIG. 5. 

in the edges and corners went unnoticed by most of the observers. 
Differences in center density were not noted very much. Variations 
in a large central portion of the entire screen were noticed and it 
was found that there is a definite field, which we call "the center 
of attention", which consists of a circle with an approximate diameter 
of 80 per cent of the height of the frame. This diameter is not 
very critical and here we could arbitrate. Again we choose standard 
inch fractions and our proposal is illustrated in Fig. 4. 

Glass cones were not used for anything larger than Leica size, 
but only the photoelectric cell itself, since the entire surface is il- 
luminated and there is no integrator required. 

130 SILENT Vol 49, No. 2 

Fig. 5 shows a diagram of the entire instrument but does not show 
the condenser system, which is located between the light source and 
the deflection mirrors. 

The procedure of transmission calibrating a lens is a very simple 
one. First, the fixed standard C* is connected to its carrier D% 
then, the variable standard Ci which is permanently connected to 
its carrier DI is set to //4. The photocells should be in balance 
now and the microammeter should show zero reading. If this is not 
the case, we have to compensate with the potentiometer until the 
instrument reads zero. Now, everything is balanced and the fixed 
standard is replaced C\ by the lens to be calibrated. 

Finally, the variable standard Cz is set to all the different / stops 
to be measured. The diaphragm control ring is turned until the 
microammeter reads zero again, and the new calibration mark is 
drawn on the control ring. The lens to be calibrated will always 
correspond to the setting of the variable standard C 2 - 

Calibrating a lens with this device can be done. by any unskilled 
person. It does not take more than about a minute to calibrate ten 
different/ stops on a lens within an accuracy of 5 per cent. 




Summary. A servo-type of mechanism has been devised which automatically will 
maintain a lens in focus under the guidance of any one of a number of distance- 
measuring devices, or which may be arranged to replace the Selsyn-type of remote 
control for focusing a lens. The mechanism requires no special cam or nonlinear 
element to fit a particular lens, but automatically solves the equation of any lens for 
which it may be adjusted at the moment. When replacing a Selsyn system this 
servomechanism has the advantage that it can never get out of step. The system is 
particularly applicable to motion picture camera lenses in follow -focus work. 

The field of photography has many opportunities for the application 
of mechanisms for focusing lenses either by remote control or through 

* Presented Apr. 21, 1947, at the SMPE Convention in Chicago. 
** Formerly Mitchell Camera Corporation, Glendale, California. 

Aug. 1947 



the action of specially cut cams or linkages. ' Frequently in follow 
shots the focus of the camera lens is varied by means of. either direct 
gearing or Selsyn motors manually operated in accordance with a 
scale calibrated to the particular lens in use and under the guidance 
of some form of distance cuing. On the optical bench a special cam 
fitted to the lens may be used to maintain focus at different distances. 
Some form of device which would at all times keep a lens in focus on 
an object as the distance to the object varies has been the purpose of 
these mechanisms. 



. 1. 

This paper describes a mechanism which departs radically from the 
types now in use, in that it will .focus any lens with which it is asso- 
ciated, regardless of focal length, without requiring special scales or 
cams to work with that lens. The operation is accomplished elec- 
trically by means of a servomechanism, may be controlled at a con- 
siderable distance from the camera without its accuracy being af- 
fected, does not get out of step, and can be made to give an accuracy 
of setting of the lens position which will meet the most exacting re- 

The similarity between the lens equation* 

* u = lens-to-object distance; v = lens-to-image distance; / = focal length 
of lens; v/u is defined as the magnification ratio. 



Vol 49, No. 2 

and that for two resistances in parallel 


r> r In = "B \") 

AI A2 J\. 

has led to the suggestion that, if two resistances be arranged in a 
Wheatstone-bridge circuit as shown in Fig. 1, and one of these resist- 
ances is varied as a function of distance, then the variation of the 
second to restore the balance of the bridge could be made to focus a 
lens. In this arrangment RI and R 2 bear a scalar relationship to the 
distances which they represent. For instance, if one inch be 






RU-F _ RF 

R =R 
FIG. 2. 

represented by 100 ohms, the resistance values R u and R v (the sub- 
scripts corresponding to the optical formula) would be 12,000 and 
203.39 ohms, respectively, for a 2-in. lens focused on an object 10 ft 
distant. With this type of bridge it is at once apparent that only a 
small percentage change in R v will be required to restore the balance 
even though a comparatively large change has been made in R u . This 
is true undermost practical operating conditions of magnification ratio, 
and has the effect of limiting the criticalness with which the lens can 
be focused. Furthermore the ratio of the gearing driving the lens re- 
sistance and the distance-measuring resistance must be such that the 
same scalar relationship exists in both. It will be seen that the adap- 
tation of this form of bridge to lenses of different focal lengths and 
mounting arrangements involves complexities not present in the form 
described below. 

Aug. 1947 



By means of a simple algebraic transformation the lens equation 
(7) can be written in the form 


Eq (3) may be represented by resistances in another form of Wheat- 
stone bridge as shown in Fig. 2, the subscripts identifying the corre- 
sponding quantities as follows : 

R u -f _ R f (4 . 

P ^ w 

K f K v -f 

As in the previous arrangement each resistance must bear a scalar 
relationship to the distance which it represents. However, the 

KR V . F 








Rp KRy-F 

FIG. 3. 

equation is explicit in vf, which is the lens displacement from its 
infinity position. Thus the resistance associated with the lens 
position in this Wheatstone bridge varies from zero to some finite 
value, a much larger percentage change than for the arrangement of 
Fig. 1. Also the full variations of R u -f and R v -f appear in the corre- 
sponding arms of the bridge of Fig. 2 without the obscuring fac- 
tor of the second resistor in parallel. This results in greater 
criticalness of focus adjustment under most working conditions. 

It should be noted at his point that Eq (3) is unchanged in its 
validity when multiplied by the constants k and m as follows : 

u - f kf 

m y-*. = m -T-. r.- (5) 

/ k(v - f) 

The constants m and k can be applied correspondingly to (4) 

134 SILENT Vol 49, No. 2 

without affecting its validity. This permits the use of different scalar 
factors for the armsof the bridge in an arrangement which will produce 
the maximum sensitivity. A practical arrangement is indicated in 
Fig. 3. In the circuit as shown the constant m has been made unity. 
By so doing, since one arm of the bridge is Rf and its adjacent arm is 
R u -f, the sum of these arms is R u and the two arms may be electri- 
cally continuous, thereby simplifying the circuit and eliminating the 
effect of contact resistance at one of the sliders. 

The equation applying to the circuit of Fig. 3 is, therefore, as 

follows : 

Ru-f _ kRf 
Rf *(*.-/)' \T ri 

Using a scalar value for R w -f and R f of 100 ohms per inch gives 
practical values for these two bridge arms. For any ordinary lens 
working at less than unity magnification a favorable arrangement is 
obtained by making k larger than unity. The maximum values of 
the circuit elements shown in Fig. 3 may be as follows: R f = 600 
ohms; R tt _/ = 60,000 ohms; kR f = 17,500 ohms; kR v - f = 500 ohms. 
Using these values in the circuit shown makes k = 42.5 when working 
with a 2-in. lens adjusted so that kR v -f = 500 ohms when the lens is 
focused at 3 ft. This constant k will vary with different lenses and is 
not a primary factor in choosing the values of the component re- 
sistors. It appears automatically at its proper value in the normal 
course of adjustment of the circuit. The bridge shown in Fig. 3, 
when working with a 2-in. lens focused at 10 ft, provides 176 times 
as much error voltage into a high-impedance galvanometer as the 
bridge shown in Fig. 1. Accordingly, the sensitivity requirement 
for the galvanometer (or a servomechanism when employed) is cor- 
respondingly reduced by this arrangement over that of Fig. 1. In 
order to have sufficient sensitivity to focus a lens having a 2-in. focal 
length to within 0.0005 in. of the exact position for an object at 10 
ft requires a sensitivity such that an adjustment will just be initiated 
when the error voltage produced by the bridge is 0.12 v with 12 v 
applied to the bridge. A sensitivity of 0.06 v is entirely practicable 
and results in an even more critical setting of the lens. 

While all four of the arms of the bridge used for the device are 
variable, only two of them, kR v -f and R u -f, the lens-position and 
distance-measuring resistors, have any particular requirement im- 
posed upon their characteristics, namely, that their variation of 
resistance shall be linear with respect to their position. Variable 

Aug. 1947 



resistors which meet this requirement within 1 per cent are now 
commercially available and are satisfactory for this service. 

Since no variation of resistors Rf and kR f is required when once 
set for a particular lens, these resistors may be linear or may be given 
any other convenient characteristic. Resistor Rf can be fitted con- 
veniently with a scale of focal lengths corresponding to the scalar 
value of distances applying to resistor R u -f. Ordinarily this cali- 
bration can be individually determined and is not at all critical, a 
variation of 10 per cent introducing negligible error in focusing the 
lens. Resistor kRf does not need a calibration scale. 

*~RF ** RU-F~*. 


FlG. 4. 

The procedure for initially setting the values of resistances in the 
Wheatstone bridge to operate with any particular lens is very simple 
and may be performed in the field in a few seconds. Referring to 
Figs. 3 and 4, the steps are as follows : First, when the lens is mounted 
on the camera, the coupling to the lens resistor kRy-f is set so that this 
resistor is on its zero position when the lens is set on its infinity-focus 
position. Second, the resistor R f is set at the value corresponding 
to the focal length of the lens, and is left at this position until 
another lens is to be used. Third, the distance resistor R u -f is set 
to the value corresponding to some convenient footage mark on 
the lens-focusing scale, usually the shortest distance for which it 
is marked. Fourth, the balancing resistor kRf is varied until the 
bridge is in balance when the lens is set at the footage mark chosen 
in the third step, as indicated by the galvanometer or by the 

136 SILENT Voi 49, No. 2 

servomotor coming to rest. The device is now in adjustment and 
the balancing resistor is left fixed at this position as long as this lens 
is in use. The correct focus of the lens will always be otained for 
any particular setting of the distance resistor when the lens is 
moved to the position which balances the bridge. 

The form in which such an electrical focusing device is con- 
structed may vary with the particular application and intended 
usage. Referring to Fig. 3, the variable resistor kR v - f should be 
mechanically associated with the lens mount in such manner that 
focusing from infinity to the closest marked setting varies the resis- 
tor from zero to approximately the maximum value, respectively. 
This variation must be linear with respect to the lens position, but 
need not have any particular ratio with respect to it. While gear 
operation between the lens and resistor is possible, a tape or short 
cable, such as used for radio-set dials, is preferable as backlash or 
lost motion is minimized. 

The other elements of the control may be mounted in a unit con- 
veniently located with respect to operation of the camera, at a dis- 
tance of several feet if desired. The distance-measuring resistor 
J R M _/may be fitted with a scale calibrated in feet and gear driven from 
either a distance-measuring tape, cyclometer, or one of the various 
types of range-finder distance-measuring devices popular in photo- 
graphic work. For instance, in a dolly shot the distance-measuring 
resistor R u -f may be adjusted by the movement of the dolly either 
through a caster running on the floor or a cable fixed to some reference 

By using a galvanometer to indicate balance and by manually ad- 
justing the lens to keep the galvanometer at zero a follow-focus 
system is possible which is superior to the present manual follow 
focus in that a precise indication of the accuracy 'of the lens setting is 
obtained and the requirement of a special scale to fit each lens is elimi- 
nated. This would be of particular value on locations where battery 
operation of the device is desirable. 

By substituting a servoamplifier and motor drive for the galva- 
nometer, the motor being arranged to focus the lens through gearing 
as indicated in Fig. 4, focusing will be made entirely automatic and 
more accurate than can ordinarily be accomplished by hand adjust- 
ment. By the use of an alternating voltage on the Wheatstone 
bridge the servoamplifier is materially simplified and may be given 
a sensitivity of 0.06 v or greater without entailing any severe 


requirements of design. The distance-measuring resistance may be 
located at any reasonable distance from the rest of the equipment, 
permitting substitution of the servodrive for Selsyn motors with the 
attendant advantages of adaptability to different lenses and the 
elimination of the synchronism requirement. Also, backlash or lost 
motion in the gear train or flexible shaft between the motor and 
the lens has no effect since it does hot appear in the positioning of 
the lens resistor kR v -f. 

In order to avoid any possible injury to the lens-driving mecha- 
nism during the lining-up process when lenses are changed and a re- 
adjustment is necessary, a slipping clutch is desirable between the 
servomotor and the lens. However, this slipping should not be 
between the lens and its positioning resistor for obvious reasons. 
With such an arrangement limit switches at the extremes of lens 
positions become unnecessary. 

The values of the circuit elements previously given for Fig. 3 (and 
which apply equally to Fig. 4) are particularly applicable to lenses of 
focal lengths from 1 to 6 in. and working at distances from 3 to 50 ft. 
By minor changes in the values of these elements the range of dis- 
tances or focal lengths accommodated can be extended as desired. 
For instance, where lenses covering a large range of focal lengths are 
to be used at near-unity magnification ratio, some advantages can be 
obtained by making the constant m (Eq 5) other than unity. 

While the device as described applies most easily to lens mounts 
such as used on motion picture cameras where rotation of the mount 
causes movement of the entire lens in translation, it may be applied to 
cameras in which the lens is moved in translation directly. 

The mechanical arrangements and electrical requirements are such 
that great flexibility in the application to a camera is possible. Ac- 
curate manual focusing under the guidance of electrical circuits con- 
trolled by a distance-measuring device represents probably the 
simplest application. By the addition of a servomotor, precise 
focusing from a remote position can be accomplished without the 
customary errors of backlash and requirements of synchronism. 
When the distance element of the electrical circuit is associated with a 
distance-measuring device the servo-operated system becomes a fully 
automatic camera-lens focusing instrument capable of accurately 
following changes in distance to an object. A simplified arrangement 
of the servosystem is adaptable to the remote focusing of a lens of a 
projector and for many other uses. 

138 SILENT Vol 49, No. 2 

Acknowledgment. The bridge circuits and servocontrol for 
focusing lenses, as shown in Figs. 2, 3, and 4, were developed for, 
and during my association with, the Mitchell Camera Corporation. 
I am indebted to Mr. Donald H. Kelley, formerly of that organi- 
zation, for the suggestion of a Wheatstone bridge employing resis- 
tances in parallel to solve the lens equation, as shown in Fig. 1. 
Applications have been filed for 'patents on these circuits and devices 
for focusing lenses. 


MR. L. L. RYDER: I think it might be well to add at this point that lens- 
focusing devices, such as this device described here by Mr. Silent and submitted 
by Mr. Lorance, are becoming more and more important as we utilize the boom 
more and more in the shooting of motion pictures and television. There is a 
definite trend under way now in the Hollywood area of motion picture production 
to lighten the boom and decrease the amount of equipment and personnel that 
are carried with the cameras. Devices of this type are accomplishing that end. 

MR. G. T. LORANCE: When I was asked to read this paper I was not familiar 
with its contents. As I read it and tried to understand it, I was really quite 
surprised by the possibilities of it as a useful tool. I am really pleased that I 
had the opportunity to see and read the paper. 

CHAIRMAN A. SHAPIRO : It appears to be a very unique solution to the problem. 

I would like to ask Mr. Lorance whether there is a time lag, however small, 
between the adjustment movement and the actual focusing? 

MR. LORANCE: I weuld guess that while there is a time lag the magnitude 
of that time lag would be a function of design and could be made so short that 
no one would ever know it. 

CHAIRMAN SHAPIRO: The paper did not have any illustrations to illustrate 
the mechanism itself. It was intimated that whatever play might exist in gearing 
would be taken up by the electrical impulses. In finding themselves in balance, 
would there be a tendency to oscillate? 

MR. LORANCE: That again is the function of the design of the servomecha- 
nism the amplifier and servomotor. ' .' ' 

Mr. Ryder has indicated that most of the movement of lens is in focusing and 
is slowly done and there probably would be no particular hazard involved in that. 

I think Fig. 4 shows the relationship involved between the servomotor and 
the lens-position device and the lens-position-measuring resistor. Slippage can 
occur. The motor can run fast or slow. The resistance is kR v - f . It is related 
so that when one moves the other does. The motor adjusts itself until it is 
balanced. That is the point which allows for a slipping clutch in the motor drive, 
if desirable. 

MR. GREEN: Someone expressed the idea that the depth of focus of the lens 
would more than mask any time lag in the motor. Is that right? 

MR. LORANCE: In reading over the paper myself I don't recall any reference 
to that. My own hunch, and it is purely a hunch, is that it would be very easy 
to make the mechanism work fast enough. I don't think that you would have 
to depend on the effect that you mentioned. 


MR. GREEN: Can anyone explain how this circuit or proposed mechanism 
would be adapted to the focusing of a projector lens or are they referring to 
background projection as used in the studio rather than theater projection? 

MR. RYDER: In Hollywood remote-control focusing equipment is used a 
great deal in background projectors. At Paramount we have a remote-control 
focusing device on our background projection equipment. The equipment that 
we use is not of the latest design. When the control dial is turned in one direc- 
tion as related to the position and brought to rest in the reverse position, it makes 
focusing very difficult, but the focusing for background projection should by all 
means be done from the camera side of the background screen. This is a device 
which should make it practical to focus definitely and definitely work to a given 
position of the control knob during the focusing on background projection work. 

MR. GREEN: The reference then was that it not be adapted to maintain a 
sharp focus on a theater screen, regardless of the buckle of the film? In other 
words, to correct the outer focus effect that comes from changing from black 
and white to color and vice versa? 

MR. RYDER: My supposition is that there is no intent that this device would 
meet that requirement. 

MR. LORANCE: I would agree with you. 

DR. E. W. KELLOGG: There was reference made to the time lag in adjusting 
the lens. I am attempting to remark that if the eye were not very tolerant of 
very brief departures from perfection there wouldn't be any Society of Motion 
Picture Engineers. 



Summary. The mass market necessarily is a price market, since only in the 
low-price brackets are sufficient purchasers found to warrant large volume production. 

Low cost can be achieved while high relative value is maintained, only if the engi- 
neering adheres carefully to these fundamentals: (a) elimination of unnecessary 
features; (b) determination of acceptable minimum levels of performance; (c) provi- 
sion for attractive external appearance; (d) design of parts for mass production by 
proper selection of dimensional tolerances, careful tooling to insure such tolerances, 
and the use of suitable materials; (e) the establishment of economical assembly-line 
procedure by considering assembly problems from the inception of parts design, 
providing accurate assembly fixtures, and maintaining adequate process inspection 
throughout manufacture. -. . . 

* Presented Apr. 25, 1947, at the SMPE Convention in Chicago. 
** Executive Vice-President, Excel Movie Products, Inc., 4234 Drummond 
Place, Chicago 39, 111. 

140 CASE Vol 49, No. 2 

Before we examine the engineering aspects of amateur projection 
equipment for the mass market, it is essential that we identify and 
isolate that market from the total demand for narrow-gauge pro- 
jection equipment. 

The mass market is composed of those who are able, or who are 
prepared, to spend a minimum sum to possess the equipment neces- 
sary to the pursuit of their hobby. It is not composed of persons 
skilled in the art or who are already earnest enthusiasts. The mass 
market includes juveniles, or their parents, and beginners who are 
either persons in modest financial circumstances or in doubt as to 
whether they will ever be deeply interested, hence only willing to 
start their adventure into home motion pictures with a small ex- 
penditure. No matter how varied their original motives may be, 
they may all be reduced to one common denominator. They are in 
search of motion picture entertainment in their homes. Their 
enjoyment of this hobby is subjective. Their satisfaction derives from 
the entertainment value of the subject matter contained in the films 
they see projected on the screen, not in pride of possession of a tech- 
nical instrument. 

To the nontechnical mind, quality is a subjective reaction expressed 
in terms of enjoyment or pleasure. To the nontechnical home motion 
picture enthusiast "the play's the thing". He is absorbed by the 
story, he is prepared to laugh at the right places in the comedy, he is 
interested in the people appearing on the screen what they do and 
how they act. He is almost unaware of the medium through which 
he is receiving his enjoyment or of the technical miracles which make 
it possible. To him a projector is of acceptable quality if the level 
of its performance does not introduce annoying distractions to his 
enjoyment. Beyond that point, mechanical or optical excellence is 
to him esoteric, unrecognizable, and unappreciated. 

When engineers witness a performance on the screen, they are in- 
clined to judge what is seen by a set of standards which are quite 
technical. Their eyes and ears have been trained to detect and dis- 
cern minute aberrations. The result is that technical aspects are seen 
which are not at all evident to the layman who is enjoying the un- 
folding of the drama and who is concerned primarily with the im- 
pressions which he receives from the story. Therefore, it is essential 
to keep in mind at all times the fundamental difference between the 
objective definition of quality as insisted upon by the technically 
trained and the subjective viewpoint held by the great mass of the 


general public which evaluates quality solely in terms of the degree 
of enjoyment which is experienced. 

The threshold level of technical quality of performance necessary 
to meet the subjective approval of persons who comprise the mass 
market has been carefully and scientifically determined. There has 
been produced, under the most exacting engineering and manufactur- 
ing controls, a projector which delivers the required degree of subjec- 
tive satisfaction at a low cost that practically every family can enjoy. 

The first consideration must be to simplify the design so as to in- 
clude only such features which have to be incorporated to maintain 
performance above the determined subjective threshold. 

Measured by such a criterion, there are numerous eliminations 
possible; for example, the framing device. Thirty-five millimeter 
films have to be framed, they can get out of frame. But why is a 
frame put on a narrow-gauge projector? A 16- or 8-mm millimeter 
film cannot be spliced out of frame. If the aperture and the claw are 
correctly interrelated, and accurately set, no framing adjustment is 
necessary. It adds to the cost, but not to the user's subjective 

A universal motor, capable of operating on either alternating or 
direct current, is an unnecessary reminder of the days gone by. It is 
superfluous today when over 99 per cent of homes are supplied with 
110-v, 60-cycle a-c. A simple a-c motor can be furnished at a much 
lower cost than a universal motor and, since it runs at a constant 
speed, 'it makes unnecessary the inclusion (at additional cost) of a 
speed-controlling rheostat. Parenthetically, many amateurs who 
use a rheostat-controlled projector inadvertently project either too 
fast or too slow, sometimes with damaging results to film or pro- 
jection lamps. With a constant-speed motor and a simple two-step 
pulley, speeds of 16 and 24 frames per sec are assured at will. It is 
only at these speeds that films should be projected. Therefore, this 
simple motor is not only an economy but a positive advantage. 

Such refinements as stop-on-film, reverse projection, and auto- 
matic rewind, though demanded by 'the advanced cinema tographer, 
do not add to the fundamental subjective enjoyment of the casual 
motion picture fan, to a degree which justifies their cost, so they are 
not included. 

A projector for the mass market must throw a clear picture on the 
screen, without ghost or flicker, without noise or clatter; in a word, 
without subjective distraction. These are essentials, but to insist 

142 CASE Vol 49, No. 2 

that the same standards of these qualities, accepted as proper for 
theater projection or for commercial or educational work, should 
apply in the mass-market home would deny home motion pictures 
to millions. 

Six or eight persons at a time constitute an average casual home 
audience. A picture of the order of 18 in. with a screen distance of 
about 9 ft is an adequate and convenient arrangement of audience, 
screen, and projector, in the mass-market livingroom. 

A small projection lamp, correctly placed and with the light beam 
well condensed, gives enjoyable projection under such conditions. 
To incorporate high-wattage illumination, with the necessary ac- 
companiment of elaborate cooling arrangements and heavier elec- 
trical components, capable of auditorium use, into the modest home 
projector, does not add to the subjective enjoyment of motion pictures 
in the livingroom sufficiently to justify the extravagant cost. As a 
matter of fact, such illumination is excessive and unpleasant under 
most home conditions. 

We must now consider a phase of the problem which has a profound 
bearing on design, but which is very difficult of exact analysis from 
the standpoint of engineering objectivity. How silent must a silent 
projector be before it is silent? This requires a careful analysis of 
each of the various types of operational noise in a projector. Here 
the engineering staff concerns itself with the matter of a practical 
balance between elaborateness of design and closeness of tolerance, 
and the significance of the noise in its subjective effect on the audi- 
ence. Mechanical noise in a projector is, of course, a question of 
degree. No projector is completely silent. Wherever it would 
require elaborate refinement of mechanism to eliminate an opera- 
tional sound, consideration is first given as to whether that par- 
ticular sound would detract from the subjective or social enjoyment 
of the picture performance under home conditions. As long as the 
noise under examination is of a quality and intensity which appears to 
be below a distracting level, we feel it is not necessary to increase the 
precision of the design or deer ease -allowable tolerances to eliminate it. 
We endeavor to make moving parts sturdy, not delicate, even though 
this may mean a sacrifice of some quietness, because we realize that 
careful maintenance of delicate parts would not be the experience of 
our mass-produced projector in the user's possession. 

Up to this point, the engineering aspects of projector design prob- 
lems have been quite subjective. They have involved consideration 


of the intended use, and the choice of design features which, when 
combined, serve best the needs which such use require. 

That these aspects are not capable of being reduced to formulas and 
graphs does not bring them outside of the realm of applied engineering 
To select correctly, and combine properly, the various components 
of a projector which will serve the mass market to give entertainment 
adequately and which, at the same time, will approach the ultimate 
practical level of economy, calls upon two of the most significant 
attributes of the engineering mind, experience and judgment. 

It is not enough, however, to have selected a projector design which 
is adequately simple. To bring about the low price level necessary 
to achieve volume, the components must be engineered so as to be 
capable of manufacture with the least expenditure of labor and 
machine time. Our engineers have joined together many components 
into one die casting, intricate, it is true, but so arranged that it is the 
entire projector except for the moving parts and for the enclosing cover. 

This one casting takes the place of 32 separate and various types 
of individually fabricated parts of a preceding design. It is so con- 
ceived that, after reaming a few holes and after painting, it starts 
down a continuously moving assembly line to have component parts 
attached to it by successive operators and, at the end of the line, is 
finally inspected, tested with a film, and then has the enclosing covers 
affixed and is packed for shipment. This assembly line, consisting of 
some 30 girls, is designed to handle one projector every 55 sec of the 
working period. For this to be a successful operation, it is obvious 
that the assembly must be conducted without hesitation or delay for 
fitting or selection of parts. This involves a high order of precision 
engineering thought and accomplishment. 

Each individual part naturally must have certain manufacturing 
tolerances. These tolerances must be so calculated and controlled 
that they will not stack up to a degree where interference or opera- 
tional gaps will be interposed. 

Long before manufacture of any part was started, the layout on 
the board was completely analyzed, each part in its relation to every 
other mating part, and to the assembly as a whole, and tolerances 
assigned, always with these two ends in view, (1) parts must be al- 
lowed as generous tolerances as possible, in order to make possible 
their fabrication by quantity-production methods, and (2) the finished 
projector must satisfy the mass-market requirement of the above- 
indicated quality of performance. 

144 CASE Vol 49, No. 2 

At certain locations in any projector, components must be held to 
very close precision. In the intermittent movement, as an example, 
the cam and the claw have to be made accurately and fitted closely. 
Our tooling engineers have developed a method of fabricating to 
very precise dimensions and closely held tolerances on these parts, 
by successive shaving operations in a punch press, keeping the limits 
down to a few ten -thousandths. The ingenuity of this method suc- 
cessfully avoids the necessity of finishing these parts on a grinder and 
yet permitting a random assembly. The resultant economies are 
importantly reflected in the eventual price. 

Certain high-speed gears are accurately generated, low-speed 
gears are die-cast, and phenolic gears are alternated with, metallic 
gears. The correct choices here make for smoothness and control 
of noise, without excessive cost. 

In short, precision is used where the end justifies the means. 
Never is precision demanded solely as an abstraction. The tooling 
engineers, the production engineer, and the chief inspector are con- 
stantly at the elbow of the designer to see to it that each part can be 
tooled adequately, can be fabricated and assembled economically, 
and can be relied upon for satisfactory use, before any drawing is 

In assembly, ordinary factory labor is used, not skilled mechanics. 
This means that jigs and fixtures must be provided to ensure that 
when the parts are placed in the chassis they are positioned correctly. 
As an example, the positioning of the shutter and its gear train to the 
cam, thence to the claw and the sprocket hole in the film, back to the 
frame of the film in the aperture, all must be exact. An ingenious 
and rather complicated fixture makes it possible for these components 
only to go together in such a way that the timing is correct. It is 
exact to the precise teeth of the intermeshed gears and the claw 
position and shutter are held to within one degree of arc of rotation. 
"We hold to accepted standards for flicker of 48 balanced interrup- 
tions per second and the cam is designed for a rapid pull-down, using, 
nevertheless, quite conventional and accepted modes of attack, 
pull, and release. 

While the screen intensity and uniformity are consistent with 
published standards, here we are constantly studying in our labora- 
tories and sponsoring considerable outside research, as well, to the 
end that the optical result may become more and more efficient. 

Here, as well as in the question of screen definition, the matter of 


the subjective impression received by the audience is a controlling 
factor. For the infrequent critical and more acute' user, more* 
elaborate projection lenses are available, and the purchase of such a 
lens as an accessory can be elected by the occasional individual, 
without penalizing the majority to whom the difference in perform- 
ance is not worth the price. 

The services of a group of industrial stylists are retained to col- 
laborate with our designers to bring to the finished product an ex- 
ternal appearance which will be not only functional, but attractive, 
and the current result of this joint effort seems to have been a for- 
tunate combination of the practical and the esthetic. 

It is seen then, that in the production of mass-market projection 
equipment, the methods are not haphazard. Engineering talent of a 
high order, not only mechanical engineering, optical, and electrical 
engineering, but, so to speak, social or human engineering, is required 
to bring about the most satisfactory final result in every respect. It 
is possible for this to be done within the cost limitations of the mass- 
produced machine where production is scheduled at a rate which can 
turn out several hundred thousand units each year, introducing a 
number of neophytes into the hobby, which means universal accep- 
tance of the efforts of scientist and industrialist alike in the exploita- 
tion of all levels of equipment, from the beautifully precise and elab- 
orate mechanisms for the professional to the simplified mass models. 

From the ranks of the hesitant beginners and the youths of today, 
learning to enjoy the marvels of home motion pictures with mass- 
produced equipment, come the experts and intense enthusiasts of 
tomorrow, graduating into the use of more elaborate products of the 
professional field. 


MR. A. SHAPIRO : What is the ratio of pull-down time of the total cycle of your 

MR. P. H. CASE: We have two designs currently being developed. One of 
them, I believe, we are working at 41 deg at the present time. That is the one 
that will be used in mass production. We have developed a pulldown that is 
capable of operation without injury to the film at 24 deg. 

MR. P. L. KARR: I might add a point to that. In one of the previous designs 
which was demonstrated here, a very large motor was demonstrated and a pull- 
down cycle of about 24 deg was indicated. . We found that a very light motor of 
perhaps one fourth to one half of the starting requirements of a universal motor 
can be used with proper design of the intermittent system so that too much work 
is not wasted in excessive pressure on the cam during the pull-down cycle. 

146 CASE 

MR. R. L. LEWIS: It would appear to me from your paper that you are classi- 
fying projector designs in five quantities. I want to see if these check: price, 
lumens, dynamic resolutions, noise, and life of the film. 

MR. CASE: That is probably a fair way of stating it, although I classify projec- 
tion equipment in two major groups, for the mass market and for the professional 
market. One is where fineness of instrument is the controlling criterion and one 
where the price is the primary consideration. Both markets exist. Both markets 
are important. It is by the existence of the first market that the second market is 
created. People are graduated from the mass market to the professional market. 

MR. RYAN: I am wondering what the area of your stockroom is compared 
to the area of your manufacture and assembly areas? 

MR. CASE: I should say that our stockrooms, raw materials, finished parts, 
and subassembly occupy a space approximately twice as large as our final assembly 
space. Of course, that does not take into consideration the fabrication of parts, 
but assembly is very streamlined. We can produce about 500 projectors on a 
single table line that is less than half as long as this room. 

DR. E. W. KELLOGG: This paper brings to my mind a comment about the 
Society that has struck me a good many times. Vast ranges of interests and points 
of view are brought together here in our meetings. There is a decided contrast in 
the utmost refinement demanded by our Hollywood producers and the low cost, 
such as this paper brings out, wanted by people who are pursuing a hobby. 

I want to make one more comment on a point that Mr. Case made, that a sound 
projector needs to be more silent than a silent projector. 

MR. CASE: That is true. I agree that there is a wide range of interests and 
activities. I was induced to prepare this paper because of my feeling that the 
emphasis should be placed in relative proportion on all types of equipment and not 
exclusively on the more technical aspects. The mass market is a large market and 
it has to be supplied. It has to be supplied efficiently if the greatest return of the 
customer's dollar is to be achieved. 



Distortion caused by speed variation in recording or reproducing 
has long been a serious source of quality degradation in sound record- 
ing. "Flutter" is the term most generally used for this type of dis- 
tortion by the motion picture industry, while the broadcast record- 
ing field knows it as "wow". 

Considerable effort has been spent in a quite successful attempt to 
eliminate, or at least reduce, the causes of this trouble to a reasonable 
minimum in motion picture equipment, but the resulting effect, which 
is often quite disturbing when audible, has resisted numerous attempts 
at definition in relatively simple but adequate terms. 

Different methods of measurements have been employed by dif- 
ferent groups, resulting in different methods of stating results, which, 
when quoted without adequate explanation, caused confusion. Some 
of the earliest methods made use of an oscillographic trace indicating 
frequency variation. The interpretation of such traces was laborious 
and required considerable skill. For service work a simpler, portable 
device was needed in which a direct meter reading of flutter or wow 
was easily obtained. Such instruments, called flutter bridges, were 
originally developed by RCA and Western Electric. Other instru- 
ments of a similar nature have since been made, but unfortunately, 
a lack of standardization makes it doubtful that useful comparisons 
of the readings between instruments can be made at present. It is 
therefore necessary to set up standard definitions for the terms to be 
used and for the quantities to be measured. 

The Sound Committee undertook to draw up a set of standard 
specifications for flutter and the first draft was submitted recently to 
the membership of the Sound Committee and also to representative 
groups in the broadcast industry for comments. Almost unanimous 
approval was obtained from the members of the Sound Committee 
and the original specifications might have been finally incorporated 

* Received July 17, 1947. 



in this report with only a few minor changes in wording. However, 
very valuable comments were received from E. W. Kellogg, of the 
Radio Corporation of America, J. L. Hathaway, of the National 
Broadcasting Company, and W. H. Offenhauser, Jr., of the Columbia 
Broadcasting System, and the original specifications have been modi- 
fied to meet several of their objections and incorporate many of their 
suggestions. The following revised draft represents a compromise 
between the original suggestions and the comments received on them 
and should be acceptable to the overwhelming majority of all con- 
cerned with this problem. These comments will be discussed later. 
The Committee presents the following revised draft as the best com- 
promise it has been able to make and hopes that it will be acceptable 
to the majority. 



This specification defines the terms to be used in describing 
the flutter or wow found in sound records and sets forth some of 
the requirements for instrumentation. 

The term "flutter" relates to any deviation of frequency 
which results, in general, from irregular motion in the recording, 
duplication, or reproduction of a tone, or from deformation of 
the record. 

NOTE 1: The term "wow" is colloquial and usually refers to deviation 
of frequency occurring at a relatively low rate as, for example, 
a "once-a-re volution" speed variation of phonograph turn- 


' J- Flutter rate is the number of excursions of frequency per 
second in a tone which has flutter. 

NOTE 1 : Each excursion is a complete cycle of deviation, for example, 
from maximum frequency to minimum frequency and back to 
maximum frequency at the rate indicated. 

NOTE 2: Flutter is usually periodic with a dominant rate. 

NOTE 3 : Two or more flutter rates may be present simultaneously each 
of which is regarded as a component of the complex variation. 

NOTE 4: Flutter which occurs at random rates close to cps is generally 
termed drift. 

Aug. 1947 




4 . 1 Per cent flutter is the ratio of the root-mean-square devia- 
tion in frequency of a tone to the average frequency ex- 
pressed as a percentage. 

NOTE 1 : Instruments which respond to the average value of frequency 
deviations or to a function intermediate between average and 
root -mean-square shall be considered satisfactory for all but 
the most critical tests, providing the indication is the root- 
mean-square value for a sinusoidal frequency variation at a 
single flutter rate. 

NOTE 2 : Readings of per cent flutter should always be accompanied by 
a statement as to the band of flutter rates wherein substantially 
uniform response is obtained. 

NOTE 3 : Per cent flutter, when measured, should not include the effect 
of any amplitude variations which may exist simultaneously 
in the tone. 

4.2 ''Per cent total flutter" is that value of flutter indicated 

by an instrument which responds uniformly to all 
flutter rates up to 200 cps. 

NOTE 1 : Instruments which respond to all rates up to 120 cps shall be 
considered adequate for all but the most critical tests. 


The standard flutter test frequency for 35-mm film and for disk 
records shall be 3000 15 cps. 

NOTE 1 : This specification is made so that test records may be employed 
interchangeably on different flutter meters. 

NOTE 2 : Use of other test frequencies is not precluded providing equiva- 
lent flutter readings can be obtained. 


Flutter index is a measure of the relative perceptibility of fre- 
quency-modulated tones. 

NOTE 1 : Flutter index I may be expressed as follows : 



where A/ = r-m-s deviation of frequency from mean in cycles 
k = per cent r-m-s flutter 
/ = frequency of tone 
r = flutter rate 

x = 1 for rates greater than 5 per second, empirical 
for lower rates 



Vol 49, No. 2 

NOTE 2: For flutter rates 1 to 5 per second the following relations are 
suggested : 

from which 

5 500' 

NOTE 3 : For flutter rates less than 1 per second the following relations 
are suggested: 

from which 




NOTE 4: The above relations are derived from flutter perceptibility 
data presented in an article by Albersheim and MacKenzie. 1 

NOTE 5: For the case where x = 1, the flutter index, when multiplied 
by \/2 is the argument of the Bessel functions of the first kind 
and the coefficients of the various orders of the Bessel func- 
tions have been shown 2 to represent the amplitudes of the cor- 
responding orders of the side frequencies present in a frequency- 
modulated tone. 

NOTE 6 : For flutter rates above 5 per second, the ear apparently hears 
the side frequencies as extraneous effects and therefore will 
perceive approximately the minimum flutter at the same value 
of flutter index over a wide range of signal frequencies, per- 
centages of flutter and rates (assuming relatively constant 
acoustic conditions). 

NOTE 7: For flutter rates less than 5 per second, the ear apparently dis- 
tinguishes the time-frequency variation rather than the dis- 
crete side frequencies and so the expression which describes 
the phenomena becomes more complicated. 

NOTE 8: The flutter index of any given device having a constant per 
^ cent flutter will vary with the signal frequency, so that the test 

frequency should always be stated with flutter index in such 
cases. Unless otherwise stated the flutter index will be as- 
sumed to refer to the standard test frequency per Section 5.0. 


The majority of the comments received indicated no objection to 
the term "flutter", but several felt that the term "wow" should be re- 
tained. To quote E. W. Kellogg (RCA), "Rather than put the term 
'wow' in parentheses as an accepted alternate to 'flutter' I suggest 


that its appropriateness, particularly for low-frequency speed varia- 
tions, be recognized in distinction to flutter as being more appropri- 
ate for rapid fluctuations but both terms are used without such limita- 
tions as, for example, 'total flutter (or wow) content.' The use of 
these words is another example of different practices having originated 
in different groups of engineers. The term 'wow' may be less digni- 
fied than flutter but was coined as an onomatopoeic term." 

A few others took a similar view. Another term which has been 
frequently used was pointed out by C. R. Daily of Paramount, who 
says : "It is suggested that the word 'drift' be referred to as indicating 
flutter rates close to cps. The word could then be excluded by a 
statement that flutter rates close to are equivalent to drift." 

The committee feels that a single term should be standard for all 
types of record frequency deviation regardless of rate. The word 
"flutter" seems more generally acceptable than "wow" for this term. 
Should usage retain the word "wow", particularly in phonograph 
work, this would not detract from the value of the definitions. 

The wording of the definition for per cent flutter was suggested by 
E. W. Kellogg. Further comments by him are as follows: "In the 
r-m-s system, over-all flutter can theoretically be predicted by taking 
the square root of the sum of the squares of the r-m-s recorder and re- 
producer flutters if they are in random relation. There is no corre- 
sponding formula for combining measurements made on peak reading 
or average reading meters. I, therefore, advocate specifications of 
flutter as an r-m-s value. This will simplify definitions, make for 
more satisfactory standardization and a more consistent relationship 
between specified flutter and the resulting quality loss." 

The chief disadvantage in specifying per cent flutter as an r-m-s 
value is that instruments to measure it are more difficult to make than 
those which respond, say, to an average value. It was felt that the 
difference between meters responding to an r-m-s and an average 
function is relatively small, so Note 1 was added to Section 4. 1 con- 
doning the use of averaging circuits for all but the most critical meas- 
urements. This compares with sound- volume-indicator practice 
wherein the calibration is for r-m-s values but the rectifying circuits 
respond more closely to the average value of incoming currents. 

Since several flutter rates are generally present, at least in film rec- 
ords, measurements of per cent flutter will usually vary with the 
bandwidth of the measuring circuits. Therefore, the definition in 
Section 4. 2 of "Per Cent Total Flutter" calls for measurement by an 


instrument which responds uniformly to flutter rates from near to 
200 cps. Some of the comments received questioned the necessity 
of setting the upper limit so high. The reason for this is that al- 
though the most common disturbance to uniform film motion is at a 
96-cps rate, the nature of the effect is such that a strong second har- 
monic usually accompanies the fundamental (caused by the sudden 
jerk as the sprocket teeth engage the film). Readings of per cent 
flutter for the band near 192 cps are frequently as high as 50 per cent 
of that for the band around 96 cps for film measured on the Western 
Electric RA-1015 Flutter-Measuring Set. 3 

However, since some instruments may not be able to pass the wide 
band mentioned and since, for most applications, this may not be a 
serious disadvantage, Note 1 was added to Section 4 . 2 to condone 
^the use of instruments which have uniform response to only 120 cps. 
It is felt that the response should extend to 120 cycles even for disk- 
flutter measurement since power-line frequencies may set up motor 
disturbances or vibrations which produce 120-cycle flutter. 

Little objection was expressed to making 3000 cps the standard 
flutter test frequency, at least for 35-mm film and for disk records. 
However, J. L. Hathaway (NBC) said : "Our present equipment em- 
ploys 4000 cps rather than 3000, since 4000 can be more easily referred 
to WWV's tone in making 'average-speed' measurements. Certain 
other measuring apparatus uses 1000 cps as standard. I wonder if it 
is necessary for apparatus to operate on a 3000-cps standard." 

It is not the purpose -of the standards being set up to preclude the 
use of any particular test frequency. However, where several test 
frequencies are employed, difficulty is bound to ensue in correlating 
the readings of different instruments and also in supplying "stand- 
ard" test records. A 3000-cycle tone was selected because, in general, 
it is simpler to make a sensitive instrument at a relatively high fre- 
quency than at a low one and also because a considerable number of 
instruments in use now employ that frequency. For measurements 
with 16-mm film, 8-mm film, and possibly some wire recording sys- 
tems, lower test frequencies may be required. 

Comment on Section 6.0 Flutter Index was the most varied of all. 
This is a new term which few have used and there was, naturally, 
considerable question as to its meaning and value. 

As anyone who has worked with the flutter problem knows, the 
per cent flutter present in a record is of little significance unless the 
rate and signal frequency of the variation are taken into consideration. 

Aug. 1947 



Thus, with a given signal frequency, roughly ten times the percentage 
of flutter is required at a rate of 100 per second to be just noticeable 
as is required at 10 per second. A term indicative of flutter percepti- 
bility without qualification as to rate would be quite valuable. 

Fortunately there appears to be a fundamental relationship between 
flutter perceptibility, rate, and signal frequency, at least for a con- 
siderable range of values. Although this relationship has not been 


10 20 

50 100 


FIG. 1 . Minimum perceptible per cent flutter for oscillator tones in small 


specifically pointed out in the literature, it is readily derived from 
published reports. 1 Curves summarizing the per cent flutter, at 
threshold for various rates and signal frequencies, which are the result 
of extensive tests at Bell Telephone Laboratories, are shown in Fig. 1 
(taken from reference 1). The per cent flutter values shown are 
0.707 times those of the original report in order to correlate with the 



Vol 49, No. 2 

r-m-s definition of per cent flutter given herein. The original data 
defined per cent flutter in terms of the peak deviation rather than the 
r-m-s values. An additional curve for 7000 cycles has been added 
from data obtained from the original tests. In Fig. 2 is plotted the 
relation of flutter index versus rate for several signal frequencies 







.- kfr ^ 



-, Kf 



100 r 






^ < 

1 i 

> ^ 


J x 


?! r 

. Q 

< *-~ 



1000 CPS TONE 
X 3000 CPS TONE 
S 7000 CPS TONE 



J 1 2 5 10 20 50 100 


FIG. 2. Minimum perceptible flutter index for oscillator tones in small 


using the data of Fig. 1 . Here the value of threshold flutter index is 
almost constant at about 0.024 between rates of 5 and 100 cps at a 
signal frequency of 1000 (where X is unity) . 'Frequencies of 3000 and 
7000 cycles show a somewhat less constant index but do not depart 
greatly. At 500 cycles the greatest departure from the constant flutter- 
index relation is obtained but this is of lesser importance since, with 
a given percentage of flutter, the condition will be most perceptible in 
higher frequencies. 

It is interesting to note that the flutter index is equal to the per cent 


flutter when the signal frequency is 100 times the rate frequency. 
Thus the two are equal for the 3000-cycle signal at a rate of 30 per 
second. This constitutes a fortunate relationship since flutter-index 
values can be considered as simply weighted per cent flutter .values 
where so desired in instrumentation. 

From the data given in Figs. 1 and 2, one can say that, at least for 
the live auditorium wherein the tests were conducted, flutter became 
perceptible when the flutter index was approximately 0.024 for all 
rates between about 5 and 100 per second and for all signal frequencies 
shown thereon. 

A reason for the above relationship is readily found by a considera- 
tion of the mathematical analysis of frequency modulation. As has 
been shown 2 - 4 - 5> 6 a frequency-modulated signal consists of a carrier 
and one or more pairs of side frequencies separated from the carrier by 
an amount equal to r, 2r, 3r, 4r, etc. The number and rela- 
tive amplitudes of these side frequencies are determined by the so- 
called modulation index a, which is defined in the literature as Kf/fm 
where K is the fractional peak change of carrier frequency and fm 
is the rate. This is similar to the flutter index per Section 6 . of the 
specification when X = 1, except that the former refers to peak-fre- 
quency changes whereas the latter relates to root-mean-square changes 
which are more suitable for instrumentation. The flutter index per 
Section 6.0 must therefore be multiplied by x/2 to equal the peak 
modulation index a, used in mathematical analysis. 

As stated in Note 5 of Section 6 . 0, the flutter index when multiplied 
by \/2 is the argument of the Bessel functions of the first kind, and 
the coefficients of the various orders have been shown to represent 
the amplitudes of the corresponding orders of side frequencies pres- 
ent. Reference to a table of Bessel functions will show that when the 
argument (flutter index multiplied by \/2) has a value of 0.03, only 
the carrier and the adjacent pair of side frequencies are significant. 
For this case, the amplitude of each side frequency is approximately 
1.5 per cent of the carrier amplitude. Since this relation holds for any 
combination of signal frequency, rate, and percentage of flutter in 
which the peak modulation index is 0.03 (r-m-s flutter index 0.021) 
it is not surprising that the ear hears roughly the same degree of 
flutter in all cases (with the exceptions pointed out in Notes 2 and 3 
of Section 6.0). In this connection, a similarity is noted to harmonic 
distortion wherein harmonic components having an amplitude be- 
tween 1 and 2 per cent of the fundamental become noticeable. 


Thus there seems to be a fundamental basis for the term ' 'Flutter 
Index", both in theory and by observation. Further investigations 
may show shortcomings and exceptions to the relationship, but the 
Committee feels that the term as defined should nevertheless prove 
a useful one and any future modifications of it may be taken care of by 
-the value assigned to the term X given in the expression of Section 6.0. 

Values for flutter index at rates below 5 cps were computed using 
the suggested values for X given in Notes 2 and 3, of Section 6.0. 
These show little change, in general, as rate and signal-frequency 
change, indicating a good approximation is obtained by the expres- 
sions for X, at least for the particular listening tests concerned. 

The qualifications given for the index for rates below 5 per second 
and referred to in Notes 2, 3, 6, and 7, of Section 6 . are, of course, 
purely empirical and subject to future evaluation. The value sug- 
gested for X for rates of 1 to 5 cycles makes the index independent of 
rate between 1 and 5 cycles, and directly proportional to per cent 
flutter. This is in agreement with general experience which shows 
that flutter perceptibility does not change greatly between about 1 
and 5 cycles (see Fig. 1). The value suggested for X for rates below 
1 per second causes the index to drop proportionally with rate below 
1 per second. This relation was proposed by J. L. Hathaway (NBC) 
who states, "We can find no evidence that the perceptibility of wow 
changes greatly between the rates of about 1 and 8 per second, and 
therefore believe that within this range the index should be a function 
of percentage of wow only. Going below 1 per second, it is apparent 
that perceptibility falls off, reaching zero for extremely slow varia- 
tions. Thus up to 1 per second the index should be a direct function 
of rate and percentage wow. Above about 8 per second it appears 
that perceptibility of wow dimimishes. 

"We propose that the wow index be defined as the perceptibility of 
wow and that, based upon present data, this should be directly 
proportional to the wow percentage and also a function of wow rate. 
The index may be expressed by 

I = Wx 

where / = wow index 

W = percentage wow 
x = 100 R when R < 1 
= 100 when 1 < R < 8 
= 800/R when R > 8 
R wow rate" 


It will be noted that the above expression takes no account of 
the signal frequency, whereas the value defined in Section 6 . does. 
If measurements are always made at the same frequency and per- 
ceptibility is in terms of that frequency, it might be satisfactory to 
ignore the signal frequency. However, the committee feels that since 
the signal frequency is an essential factor in the perceptibility of flut- 
ter tones it should be included in the expression for flutter index. 
Obviously, if the flutter index for any particular record is stated, that 
value can be true of only one signal frequency, and it would seem 
reasonable to express it for frequency of measurement, namely, the 
standard 3000 cycles. For lower frequencies the index would fall off 
and for higher frequencies it would be proportionally higher. For this 
reason a measured value of flutter index must be stated with the cor- 
responding signal frequency as, for example, "3000-cycle flutter 

The fact that flutter index, varies with frequency for a given record 
or machine is reasonable since with a given per cent flutter and rate 
the higher frequencies will appear to have the most flutter and at the 
lower frequencies the flutter may seem to disappear completely. 

Hathaway's recommendation would place the transition point 
where X ceases to be unity at 8 cycles instead of 5 as outlined in 6.0. 
This, however, is a rather small difference. The committee feels that 
the data analyzed 1 to obtain the figure used herein is probably some- 
what more extensive than that used by Hathaway in arriving at his 
conclusion. The exact point of transition is not thought to be vital 
in any event and the figures presented by Hathaway should be con- 
sidered reasonably close to those presented herein. 

The question has been raised as to what use there is for the term 
' 'Flutter Index' ' . The most important one is that it provides the user 
with a scale with which to judge the relative perceptibility of a given 
flutter condition. The per cent flutter reading does not give him 
such a measure unless weighed against rate. A second reason is that 
instruments will probably be made eventually to measure flutter index 
instead of per cent flutter. 

Kellogg has proposed a somewhat different concept for a flutter 
index as follows : 

"I think it cannot be very far from the truth to assume that the 
magnitude of the damage to quality is proportional to the square of 
the speed fluctuation. The fact that there is a threshold of objec- 
tionable wow which varies from one person to another and varies with 


the type of reproduced sound, and that above the threshold the im- 
pairment seems to increase rapidly, indicates that the square law is 
probably not a bad approximation." 

The flutter index as defined does not appear to convey such a rela- 
tion, although it does not contradict it. Actually, the flutter index 
is an expression for relative perceptibility at the threshold point 
Interest is generally centered in whether a given amount of flutter i 
noticeable rather than in just how bad it is, beyond the point of per- 

Adoption of the recommendations in this specification will, it is 
believed, serve to co-ordinate measurements and standardize instru- 
ments in a field of increasing importance. 

The chairman of your Committee wishes to express his appreciation 
to the members for their invaluable assistance and particularly tc 
R. R. Scoville without whose help and guidance this report would no1 
have been possible. 

JOHN G. FRAYNE, Chairman 







1 ALBERSHEIM, W. J., AND MACKENZIE, D. : "Analysis of Sound Film Drives", 
/. Soc. Mot. Pict. Eng., XXXVII, 5 (Nov. 1941), p. 452. 

2 SHEA, T. E., MACNAIR, W. A., AND SUBRIZI, V.: "Flutter in Sound Rec- 
ords", /. Soc. Mot. Pict. Eng., XXV, 5 (Nov. 1935), p. 403. 

3 SCOVILLE, R. R.: "A Laboratory Flutter Measuring Instrument", /. Soc. 
Mot. Pict. Eng., XXIX, 2 (Aug. 1937), p. 209. 

* 4 RODER, H. : "Amplitude, Phase, and Frequency Modulation", Proc. I.R.E., 
19, 12 (Dec. 1931), p. 2145. 

6 VAN DER POL, B.: "Frequency Modulation", Proc. I,R.E., 18, 7 (July, 
1930), p. 1194. 

6 WATSON, G. N.: "Theory of Bessel Functions", Cambridge University Press, 
1944, Macmillan Co., New York, N. Y. 

7 SHOWER, E. G., AND BIDDULPH, R.: "Differential Pitch Sensitivity of the 
Ear", /. Acous. Soc. Amer., 3, 2 (Oct. 1931), p. 275. 

8 CARSON, J. R.: "Notes on the Theory of Modulation", Proc. I.R.E., 10, 1 
(Feb. 1922), p. 57. 


9 KELLOGG, E. W.: "A New Recorder for Variable Area Recording", /. Soc. 
Mot. Pict. Eng., XV, 5 (Nov. 1930), p. 653. 

10 BELAR, H., AND KELLOGG, E. W.: "Analysis of Distortion from Sprocket 
Hole Modulation", /. Soc. Mot. Pict. Eng., XXV, 6 (Dec. 1935), p. 492. 

11 KELLOGG, E. W., AND MORGAN, A. R. : "Measurement of Speed Fluctuations 
in Sound Recording and Reproducing Equipment", /. Acous. Soc. Amer., (April 

1936), p. 271. 

12 KELLOGG, E. W.: "Review of the Quest for Constant Speed", J. Soc. Mot. 
Pict. Eng., XXVIII, 4 (April 1937), p. 337. 

. 13 COOK, E. D.: "The Technical Aspect of the High Fidelity Reproducer", 
/. Soc. Mot. Pict. Eng., XXV, 4 (Oct. 1935), p. 289. 

14 SCOVILLE, R. R.: "A Portable Flutter Measuring Instrument", /. Soc. 
. 'Mot. Pict. Eng., XXV, 5 (Nov. 1935), p. 416. 

15 LOOMIS, F. J., AND REYNOLDS, E. W. : "A New High Fidelity Sound Head", 
/. Soc. Mot. Pict. Eng., XXV, 5 (Nov. 1935), p. 449. 

16 CHINN, H. A.: "Glossary of Disk Recording Terms", Proc. I.R.E., 33, 7 
(Nov. 1945), p. 760. 

17 ROYS, H. E.: "The Measurement of Transcription Turntable Speed Varia- 
tion", Proc. I.R.E., 31, 2 (Feb. 1943), p. 52. 

18 MINER, C. R.: "Wow Meter", Gen. Elec. Rev., 47, (April 1944). 

19 FURST, V. R. : "Periodic Variations of Pitch in Saund Reproduction by 
Phonographs", Proc. I.R.E., 34, 11 (Nov. 1946), p. 887. 

20 SEAR, A. W. : "Wire Recorder Wow", /. Acous. Soc. Amer., 19, 1 (Jan. 1947), 
p. 172. 

The following references deal mainly with mechanical problems of 
flutter reduction : 

21 BLATTNER, D. G., AND ELMER, L. A.: "Machine for Cutting Master Disk 
Records", /. Soc. Mot. Pict. Eng., XIII, 37 (May 1929), p. 227. 

22 STOLLER, H. M.: "Synchronization and Speed Control of Sound Pictures", 
/. Soc. Mot. Pict. Eng., XII, 35 (Sept. 1928), p. 696. 

23 P'FANNENSTIEHL, H.: "A Reproducing Machine for Picture and Sound", /. 
Sdc. Mot. Pict. Eng., XIII, 38 (May 1929), p. 253. 

24 DREW, R. O., AND KELLOGG, E. W. : "Filter Factors of the Magnetic Drive", 
/. Soc. Mot. Pict. Eng., XXV, (Aug. 1940), p. 403. 

25 CHANDLER, J. S. : "Some Theoretical Considerations for the Design of 
Sprockets for Continuous Film Movement", /. Soc. Mot. Pict. Eng., XXXVII, 2 
(Aug. 1941), p. 164. 



Industry-wide agreement on a standard method of measuring 
flutter in 35-mm motion picture sound film reproducers is now being 
sought by the motion picture industry through the Society of Motion 
Picture Engineers and the Research Council of the Academy of Mo- 
tion Picture Arts and Sciences. The ultimate success of the project 
depends primarily upon industry acceptance of the basic standards. 

A parallel report of the Sound Committee, entitled "Proposed 
Standard Specifications for Flutter or Wow in Sound Records", At- 
tempts to set up standard definitions for the terms used in flutter 
analysis, and to standardize certain characteristics of instruments re- 
quired to measure flutter. This should place sufficient requirements 
on instruments to ensure that measurements made by different persons 
on different apparatus will give the same results. 

This report describes the characteristics of a substantially flutter- 
free test film, of known and controlled characteristics, that can be pro- 
duced practically and that will be available to serve as a reliable and 
uniform source of test signals. The proposal is presented as a form 
of specification stipulating the essential characteristics of the finished 
film, but not restricting the type of sound track, the recording equip- 
ment, or the type of film used. 

Pertinent references to previously published material on the flutter 
question were included in the report mentioned above and for that 
reason are not repeated here. 



1 . 1 This specification defines the general characteristics of. 
test films suitable for flutter measurements upon 35-mm 
motion picture film reproducers. 

2 . 1 The test film shall carry a sound track meeting standard 
release specifications with either variable-density or vari- 
able-area modulation. 

* Received July 17, 1947. 


2 . 2 The test film shall be an original record processed as out- 
lined in Section 4 . 0. 

3.1 The signal frequency shall be 3000 15 cps, sinusoidal, 
when run at a standard speed of 90 feet per minute. 

3.2 The signal frequency shall be recorded at an amplitude 
level of approximately 80 per cent of the overload value 
of the modulator or medium. 


4 . 1 Variable-area records shall be processed in a manner which 
produces minimum cross modulation according to the 
standard test 1 for this type of record when reproduced as 
a negative. 

4.2 Variable-density records shall be processed as high -con- 
trast toe negatives in a manner which produces minimum 
intermodulation according to the standard test for this 
type of record. 2 

NOTB 1 : Variable-density films should be processed in such a manner 
that sprocket-hole modulation, due to uneven development, 
will be at least 30 decibels below full modulation. 

NOTE 2: The signal-to-noise ratio of the test film shall be at least 30 
decibels when reproduced in the normal manner. 


5 . 1 The flutter content of the test film shall be specified by 
the supplier. The figure or figures furnished shall be as 
obtained when the film is reproduced on a device which is 
relatively free of flutter and measured on a flutter-measur- 
ing instrument which conforms to the specifications given 
in "Proposed Standard Specifications for Flutter or Wow 
as Related to Sound Records". 

5.2 The "Per Cent Total Flutter" shall be a minimum and 
shall not exceed 0.06 per cent. 

5 . 3 The "Per Cent Flutter' ' at any rate within an octave band 
shall not exceed 0.035 per cent. 


1 BAKER, J. O., AND ROBINSON, D. H. : "Modulated High Frequency Recording 
as a Means of Determining Conditions for Optimum Processing", /. Soc. Mot. 
Pict. Eng., XXX, 1 (Jan. 1938), p. 3. 

1 FRAYNE, J. G., AND SCOVILLE, R. R. : "Analysis and Measurement of Distor- 
tion in Variable Density Recording", J. Soc. Mot. Pict. Eng., XXXII, 6 (June 
1939), pp. 648-672. 

Catalog of 


Motion Picture Research Council, Inc. 

1421 North Western Avenue 
Hollywood 27, California 

Society of Motion Picture Engineers 

Hotel Pennsylvania 
New York i, New York 


The 35-mm and 16-mm test films listed in this catalog are made 
available by the Motion Picture Research Council and the Society of 
Motion Picture Engineers to be of more service to the motion picture 
industry in general and to the exhibitor in particular. 

Prices include shipping charges to all points within the United States. 

All test film is sold on a cost basis. Therefore, no cash discounts are 
given and facilities for extending credit are not available. 


Test Film 


(in Feet) 


35-Mm Visual Test Film 
Focus-and- Alignment Section 
Travel-Ghost Target Section 
Jump-and- Weave Target Section 



$ 17.50 

35-Mm Theater Sound Test Film 




35-Mm Multifrequency Test Film 
Type A Laboratory Type 
Type B Service Type 




35-Mm Transmission Test Film 




35-Mm Buzz-Track Test Film 


50 min* 


35-Mm Scanning-Beam Illumination 
Test Film 
Type A 17 Position Track 
Type B Snake Track 




35-Mm Sound-Focusing Test Film 
Type A 9000-Cycle Track 
Type B 7000-Cycle Track (Area) 
Type C 7000-Cycle Track (Den- 
Type C Acetate Base 



50 min 
50 min 

50 min 
50 min 



35-Mm 3000-Cycle Flutter Test Film 


50 min 


35-Mm 1000-Cycle Balancing Test 
For Two Machines 
For Three Machines 
1000-Cycle Test Film 


50 min 


35-Mm Multifrequency Warble Test 




16-Mm Sound-Projector Test Film 




16-Mm Multifrequency Test Film 




16-Mm Buzz-Track Test Film 




16-Mm Scanning-Beam Illumination 
Test Film 
Laboratory Type 
Service Type 




16-Mm Sound-Focusing Test Film 
Laboratory Type 
Service Type 




16-Mm 3000-Cycle Flutter Test Film 




16-Mm 400-Cycle Signal-Level Test 








(in Feet) 


35-Mm Visual Test Film VTF-1 450 $17.50 

The Visual Test Film is a print on safety stock, picture only containing four 
targets to check focus and alignment, travel ghost, jump and weave, and lens 
aberration. This test film is used when installing new projectors and screens or 
performing maintenance operation on existing equipment. 

The Focus-and-Alignment target shows whether or not picture size and screen 
masking are correct, and whether the projected picture is centered properly on the 

The Travel-Ghost target shows improper timing of the shutter quite readily 
and gives a clear indication of the correct adjustment as the timing is being cor- 

The Jump -and- Weave target gives an accurate indication of the unsteadiness 
of the projected picture. Picture jump is measured in per cent of picture height, 
and picture weave is measured in per cent of picture width. 

The Lens-Aberration target shows picture distortion and gives an indication of 
the lack of sharpness that will be present in pictures shown on any particular 

Explanatory titles precede each section and an instruction booklet is furnished 
giving complete details on its proper use. 

Because some users prefer loops or continuous lengths of the separate target 
sections for adjusting machines, one at a time or in pairs, separate sections of the 
first three targets have been made available. They may be purchased separately 
in 100- to 900-ft lengths in multiples of 100 ft. 

Focus-and-Alignment Section VTF-FAS 100 $5.00 

Travel-Ghost Target Section VTF-TGS 100 ,5.00 

Jump-and-Weave Target Section VTF-JWS 100 5.00 

35-Mm Theater Sound Test Film ASTR-3 500 $17.50 

The Theater Sound Test Film is a print on nitrate base containing picture and 
sound and is used to check the over-all sound quality in the theater. Included are 
main title music to check the frequency range and the high- and low-frequency 
balance, specially selected samples of current dialog recording to check the fre- 
quency response, and a piano recording to check flutter. 

Standard electrical characteristics for the commonly used types of two-way 
theater reproducing equipment were specified by the Research Council early in 
1937 in order that studios might re-record for the best possible reproduction on a 
theater sound system set to a standard electrical characteristic applicable to that 




(in Feet) 


system. The standard electrical characteristics were arrived at after listening 
tests were conducted in various representative theaters. During these tests the 
equipment in each theater was adjusted to various settings of the electrical 
characteristic. That setting which gave the optimum reproduction was estab- 
lished as the standard for that particular loudspeaker system. For the listening 
tests necessary in arriving at the standard electrical characteristics, the Committee 
devised a test film containing sample recordings of music and dialog from all the 
studios. The use of this film was so successful that prints subsequently were 
made available for the use of theater service engineers. 

This test film has been revised from time to time, in order to increase its value 
for theater service engineering use. 

The current version, designated Theater Sound Test Film ASTR-3, contains 
three dialog samples, a choral-music sample, a vocal (single-voice) music sample, 
and a sound-effect sample, totaling approximately 500 ft in length. A title is 
superimposed over the picture indicating the particular sound difficulty which 
that sample demonstrates. The material contained in the film is not necessarily 
the best recording available, but each sample has been selected to demonstrate a 
particular point of difficulty in the adjustment of theater sound systems. 

35-Mm Multifrequency Test Film 

Type .4 Laboratory Type APFA-1 450 . $25.00 

Type B Service Type ASFA-1 300 17.50 

The Multifrequency Test Film is a variable-area print on nitrate base and is 
used to obtain the electrical frequency response at the output of the power 
amplifier. Each print is individually calibrated, and correction factors, accurate 
to within - 0.25 db, are provided with each film. The response within any one 
frequency will vary less than it 0.25 db. 

Type A (Laboratory Type), normally used by manufacturers and in the instal- 
lation of equipment contains the following frequencies each, preceded by a spoken 
announcement : 

cps cps cps cps 

1000 200 1500 6000 

40 300 2000 7000 

55 400 2500 8000 

70 500 3000 9000 

100 700 4000 10000 

150 1000 5000 1000 

Type B (Service Type), normally used in routine theater servicing, includes 
the following frequencies, each preceded by a spoken announcement: 

cps cps cps cps 

1000 300 2500 5000 

40 500 3000 6000 

70 1000 3500 7000 

100 2000 4000 8000 

166 CATALOG OF TEST FILMS Vol 49, No. 2 



(in Feet) 


35-Mm Transmission Test Film TA-1 - 250 $17.50 

The Transmission Test Film is a print on safety stock and is used to check the 
position of the scanning beam, the electrical frequency response of the system, 
and flutter. 

The film contains two sound tracks, one on each side of the film, which are 
printed head to tail thereby making rewinding unnecessary when playing both 

The first track, entitled multifrequency test, starts with a short section of buzz 
track to check the guide-roller adjustment. Next is a series of the following con- 
stant frequencies: 1000, 40, 70, 130, 300, 500, 2000, 3000, 7000, and 8000 cps. 
The low frequencies are recorded at a reduced level and announcements precede 
all sections up to and including the 3000-cycle section. An unmodulated track 
of average density is included between the 7000- and 8000-cycle tones to evaluate 
the effect of film noise on the readings. 

The second track, on the opposite side of the film, entitled flutter test, contains 
60 ft of sustained piano chords followed by a 3000-cycle tone. While the flutter 
in this print is low, it is intended only as an aural check. For the bridge-type of 
flutter measurement, the toe-recorded variable-density negative (A3KC-1) should 
be used. 
35-Mm Buzz-Track Test Film ABZT-1 50 min $0 . 04/f t 

The Buzz-Track Test Film is a print on nitrate base and is used for checking 
scanning-beam placement. The track consists of an 0.087-in. opaque center with 
a frequency of 300 cycles on the picture side and a frequency of 1000 cycles on the 
sprocket side. These tracks are accurately located on the film so that when the 
film is run on a projector in correct adjustment and free from weave, no sound is 
heard. If the scanning-light beam is out of adjustment laterally, either the 300- 
or the 1000-cycle tone will be heard. 

This film is available in 50- to 500-ft lengths in multiples of 50 ft. 
35-Mm Scanning-Beani Illumination 
Test Film 

Type ,4 17-Position Track A17P-1 230 $12.50 

Type B Snake Track AST8-1 8 0.50 

The Scanning-Beam Illumination Test Film is a print on nitrate base and is used 
to check the uniformity of illumination across the scanning slit. 

Type A (17-Position Track) is used by manufacturers or on new installations. 
The film contains 17 incremental 1000-cycle tracks, all with the same amplitude 
of approximately 0.007 in. The tracks appear on the film in succession, each pre- 
ceded by an announcement identifying the track number. The 17 tracks cover 
a width greater than the standard scanning beam. By running this test film and 
observing the indications of the output meter it is possible to correct unevenness 
of illumination and bring the variation of output within a limit of Z 1.5 db, which 
is the recommended maximum variation. This is accomplished by adjusting or 
replacing the exciter lamp. 

A calibration sheet giving the exact position of each track from the guided edge 
is provided with each film. 



Type B (Snake Track) is used as an 8-ft loop for quick service adjustment of 
the scanning-beam illumination. It contains a 1000-cycle track with a 0.007-in. 
amplitude placed on the film in such a way that the track moves across the scan- 
ning slit from one edge to the other at a uniform rate. 

In order to maintain a constant length of track, and thus hold the scanned area 
constant, the usual type of film splice should not be used in making up this loop. 
Instead, a butt-end splice should be employed, obtained by placing the ends of the 
print securely against" one another without any overlap and joining the two ends 
with transparent tape, such as Scotch cellophane tape. Experience has shown this 
splice to be very practicable as it may be remade without any loss of film. In addi- 
tion, this type of splice disturbs the reading of the volume indicator less than the 
conventional overlap splice. This film has been prepared for testing the uni- 
formity of the illumination across the scanning slit and is not intended for use to 
determine slit placement adjustment, for which the Buzz Track should be used. 
35-Mm Sound-Focusing Test Film 

Type ,4 9000-Cycle Track A9KC-1 50 min $0.035/ft 

Type B 7000-Cycle Track (Area) . A7KC-1 50 min 0.035/ft 

Type C 7000-Cycle Track (Density) D7KC-1 50 min 0.035/ft 

Type C Acetate Base D7KCS-1 50 min 0.004/ft 

The Sound-Focusing Test Films are prints on nitrate base (except D7KCS-1 
on acetate base) and are used to adjust the focus and azimuth of soundhead 
optical systems. 

Type A 9000-Cycle Track (A9KC-1) contains a 9000-cycle variable-area 
tone recorded at 1 db below 100 per cent modulation with a power output varia- 
tion of less than it 0.25 db. This film is normally used by manufacturers and 
laboratories. It is not recommended for theater use. 

Type B 7000-Cycle Track (A7KC-1) contains a 7000-cycle variable-area 
tone recorded at 2 db below 100 per cent modulation with a power output varia- 
tion of less than i 0.25 db. This film is used for servicing theater equipment. 

Type C 7000-Cycle Track (D7KC-1} contains a 7000-cycle variable-density 
tone recorded at 2 db below 100 per cent modulation with a power output varia- 
tion of less than i 0,25 db. This film is used for servicing theater equipment. 
Type C is also available on acetate base (D7KCS-1}. 
These films are available in 50- to 200-ft lengths in multiples of 50 ft. 
35-Mm 3000-Cycle Flutter Test Film A3KC-1 50 min $0.05/ft 

The 3000-Cycle Flutter Test Film is a toe-recorded variable-density negative 
on nitrate base and is used in measuring flutter. A flutter bridge is required to 
make this measurement. The total flutter of this film is not more than 0.06 per 
cent. A complete analysis of the flutter content is furnished with each purchase 
of film. 

This film is available in 50- to 1000-ft lengths in multiples of 50 ft. 
35-Mm 1000-Cycle Balancing Test Film 

For Two Machines ABL2-1 14 $0.50 

For Three Machines ABL3-1 21 0.75 

1000-Cycle Test Film ABLN-1 50 min 0.035/ft 



Vol 49, No. 2 




(in Feet) 


The 1000-Cycle Balancing Film is a print on nitrate base and is used as a loop 
to measure and adjust the power -level output of two or more projection machines. 
It contains a 1000-cycle variable-area tone with a power-level output variation of 
less than 0.25 db. 

The ABL2-1 contains sufficient film for making loops for two machines and the 
ABL3-1 contains sufficient film to make loops for three machines. An instruction 
booklet is furnished with these balancing loops. 

This film is also available in single lengths of 50 to 200 ft in multiples of 50 ft. 

35-Mm Multifrequency Warble Test 

Film APWA-1 450 $25.00 

The Multifrequency Warble Test Film is a variable-area print on nitrate base 
and is used to make acoustical-response measurements. This measurement re- 
quires the use of a sound-level meter. Each print is individually calibrated and 
correction factors are provided with each film. This film contains the following 
frequencies with the indicated amount and rate of warble for each frequency: 


Amount of 

Rate of 


Amount of 

Rate of 



5 cps 


_!_ 12^/0^ 



= fc 12 1 / 2 % 



-- 1 2^/n^ 









= t -12 1 /2% 






12 1 / 2 % 

2 x / 2 cps 











= =12 1 / 2 % 






= fc !2 1 /2% 

3 cps 





= fc 12 1 / 2 % 

3 cps 


=fc 125cps 



12 1 /2% 

3 cps 





= t 12 1 / 2 % 

3 cps 





= fc !2 1 /2% 

4 cps 





= i =12 1 /2% 

4 cps 






d =12 1 /2% 

= fc 12 1 /*% 

4 cps 
4 cps 
4 cps 
4 cps 





=fc 1 2*7 ^ 

5 cps 


=t 125cps 



=t 1 2^-7 ^7 

5 cps 





=fc 1 2^-7 ^7 

5 cps 





12V 2 % 

5 cps 

* An identifying beat tone precedes these frequencies. 



16-Mm Sound-Projector Test Film Z52.2 

The 16-Mm Sound-Projector Test Film is a print on safety base, containing 
picture and sound, and is used to check the adjustment of 16-mm sound motion 
picture projection equipment and to judge the acoustics of the room in which 



the equipment is operated. This film is the 16-mm version of the 35-mm 
Theater Sound Test Film (ASTR-3) and contains the same main title and choral 
music, dialog samples, and piano recording. The picture has been optically re- 
duced from 35 to 16 mm and the sound re-recorded from 35 to 16 mm. 
16-Mm Multifrequency Test Film Z22.44 150 $41.25 

The Multifrequency Test Film is a direct-positive original recording on safety 
base and is used to obtain the electrical-frequency response at the output of the 
power amplifier. Each film is individually calibrated on equipment correct with- 
in 0.25 db up through 3000 cycles and within 0.5 db above 3000 and through 
7000 cycles. The deviation from the intended flat-response characteristic (as- 
suming negligible reproducing light-beam width) is stated as a correction for 
each frequency which will give the true level when it is added algebraically to the 
output-level measurement obtained when using the film. 

This test film contains the following series of frequencies, each preceded by a 
spoken announcement: 

cps cps cps cps 

400 300 2000 5000 

50 500 3000 6000 

100 1000 4000 7000 

200 400 

16-Mm Buzz-Track Test Film Z52.10 100 $27.50 

The Buzz-Track Test Film is an original negative on safety base and is used for 
checking scanning-beam placement. The track consists of an 0.076-in. opaque 
center with a frequency of 300 cycles on the picture side and a frequency of 1000 
cycles on the sprocket side. These tracks are accurately located on the film so 
that when the film is run on a projector in correct adjustment and free from weave, 
no sound is heard. Either or both the 1000- and 300-cycle tones will be heard, 
however, if the scanning-light beam is out of position. 

16-Mm Scanning-Beam Illumination 

Test Film 

Laboratory Type Z52.7-L 100 $27.50 

Service Type Z52.7-S 100 27.50 

The Scanning-Beam Illumination Test Film is a print on safety base and carries 
a narrow sound track (0.005 in. wide) modulated at constant level by a 1000-cycle 
tone. The location of this sound track changes at a uniform rate along the length 
of the film from a position just inside one edge of the scanned area to a position 
just inside the opposite edge of the scanned area. The narrow 1000-cycle sound 
track sweeps across the scanning-light beam from one end to the other at a uni- 
form rate, the position of the sound track relative to the ends of the light beam at 
any instant being shown by an animated diagram appearing in the picture area. 

If the scanning-beam illumination were absolutely uniform across the width 
of the scanned area, the output level of the 1000-cycle tone would be constant. 
In practice, however, some variation of an output-meter reading will always be 



observed. By running a loop of the film continuously and observing the indica- 
tions of the output meter while adjustments are made, it is usually possible to 
correct unevenness of illumination and bring the variation of output within a 
limit of l.Sdb. 

The Laboratory Type may be spliced into 34-ft loops and the Service Type may 
be spliced into 3 1 /2-ft loops. Each type is available in 100-ft lengths. 

16-Mm Sound-Focusing Test Film 

Laboratory Type Z22. 42-7000 100 $27.50 

Service Type Z22. 42-5000 100 27.50 

The Sound-Focusing Test Film is an original negative on safety base and carries 
a special "square- wave" track, chosen because its output changes more rapidly 
with changes in the focus of the sound optical system of the projector than the 
output from the usual "sine-wave" high-frequency track. The "square-wave" 
track also gives a more sensitive indication of the errors of the "azimuth" adjust- 
ment of the sound -reproducing light beam. 

The Sound-Focusing Test Film is made in two types : Laboratory Type, a 7000- 
cycle record for manufacturing and precision adjustment of the focus and azi- 
muth of the sound optical system, and Service Type, a 5000-cycle record for 
quick service adjustment. 

16-Mm 3000-Cycle Flutter Test Film Z22 .43 380 $104.50 

The 3000-Cycle Flutter Test Film is a direct-positive original recording on safety 
base and carries a 3000-cycle tone having extremely low flutter content for use 
in measuring the flutter introduced by 16-mm sound reproducers. The recorded 
frequency is within 25 cycles of the 3000-cycle frequency, the output level is 
constant within 0.25 db, and the total flutter content of the film at the time of 
shipment is less than 0.1 per cent. 

16-Mm 400-Cycle Signal-Level Test Film Z22 .45 100 $27.50 

The 400-Cycle Signal-Level Test Film is a direct-positive original recording de- 
signed to furnish as nearly as is practicable an absolute standard of recorded signal 
level'for use in measuring the effective amplification and sound output of 16-mm 
sound motion picture projectors, taking into account the sound optical system 
and phototube, as well as the amplifier and loudspeaker. 

A definite output level is determined by specifying the amplitude of the recorded 
signal, the density of the image, and the combined base and fog density of the 
clear part of the sound track within narrow limits. The specified level is approxi- 
mately 2 db below the maximum level possible and is about equal to the highest 
level that is to be expected in most recording, since in commercial practice the 
image density is usually not so great and the fog density not so low as the values 
specified for this film. 

The actual measured values of signal amplitude, image density, and fog density 
are given with each film, together with the corresponding calculated value of over- 
all deviation from the intended standard signal level. 



The following six American Standards on Motion Pictures have been approved 
recently by the American Standards Association. . 

The Standard, "Dimensions for 16-Tooth 35-Mm Motion Picture Projector 
Sprockets", Z22.35-1947, is a revision of Z22.35-1930. It will be noticed that 
the revised Standard includes a change in the B dimension from 0.945 in. to 0.943 
i 0.001 in. This change was recommended by the Standards Subcommittee 
on 35-Mm Sprockets after tests of film and sprocket life were conducted, using 
sprockets with three different B dimensions, in several theaters. 

The four standards on cutting and perforating, Z22.5-1947, Z22. 12-1947, 
Z22. 17-1947, and Z22.36-1947 are also revisions of older standards. The dimen- 
sioning technique has been changed to conform with actual methods of measure- 
ment, that is, from edge to edge of the hole rather than from center to center. 
In addition, the tolerances and form have been brought up to date. 

The Standard on "Nomenclature for Motion Picture Film Used in Studios and 
Processing Laboratories", Z22.56-1947 was formerly approved as American War 
Standard Z52J4-1944. This Standard was printed in the April, 1945, JOURNAL 
and since the only change involved is that of the foreword, it is not being reprinted 

Copies of these six standards and several more, which will be published in the 
JOURNAL shortly, will be available from the General Office of the Society in the 
very near future. It is planned to distribute these standards on S l / 2 - X 11-in. 
sheets, punched to fit the SMPE Standards binder. 


Z22.5 -1947 Cutting and Perforating Dimensions for 16-Mm Silent Motion 
Picture Negative and Positive Raw Stock 

Z22. 12-1947 Cutting and Perforating Dimensions for 16-Mm Sound Motion 
Picture Negative and Positive Raw Stock 

Z22. 17-1947 Cutting and Perforating Dimensions for 8-Mm Motion Picture 
Negative and Positive Raw Stock 

Z22. 35-1947 Dimensions for 16-Tooth 35-Mm Motion Picture Projector 

Z22.36-1947 Cutting and Perforating Dimensions for 35-Mm Motion Picture 
Positive Raw Stock 

Z 2 2. 56-1947 Nomenclature for Motion Picture Film Used in Studios and Proc- 
essing Laboratories 




Vol 49, No. 2 

American Standard 
Cutting and Perforating Dimensions for 

16-Millimeter Silent Motion Picture 

Negative and Positive Raw Stock 

R,f . V. S. Pal. Off. 

Z 2 2.5- 1947 
Revision of 


Page I of 2 pages 







0.629 0.001 

15.98 0.03 


0.3000 0.0005 

7.620 0.013 


0.0720 0.0004 

1.83 0.01 


0.0500 0.0004 

1.27 0.01 


0.036 0.002 

0.91 0.05 


Not > 0.001 

Not > 0.025 


0.413 0.001 

10.490 0.025 


30.00 0.03 

762.00 0.76 




These dimensions and tolerances apply to the material immediately after 

cutting and perforating. 

*ln any group of four consecutive perforations, the maximum difference of 

pitch shall not exceed 0.001 inch and should be as much smaller as possible. 

(This requirement has been added to the previous standard Z22.5-1941.) 
tThis dimension and tolerance was given in respect to the center line of the 

perforations in the previous standard Z22.5-1941. 
JThis dimension represents the length of any 100 consecutive perforation 



American Standard 
Cutting and Perforating Dimensions for 

16-Millimeter Silent Motion Picture 

Negative and Positive Raw Stock 

Rrg. V. S. Pal. Off. 

Revision of 

Page 2 of 2 pages 


The dimensions given in this standard represent the practice of film manu- 
facturers in that the dimensions and tolerances are for film immediately after 
perforation. The punches and dies themselves are made to tolerances con- 
siderably smaller than those given, but owing to the fact that film is a plastic 
material, the dimensions of the slit and perforated film never agree exactly 
with the dimensions of the punches and dies. Shrinkage of the film, due to 
change in moisture content or loss of residual solvents, invariably results in 
a change in these dimensions during the life of the film. This change is 
generally uniform throughout the roll. 

The uniformity of perforation is one of the most important of the variables 
affecting steadiness of projection. 

Variations in pitch from roll to roll are of little significance compared to 
variations from one sprocket hole to the next. Actually, it is the maximum 
variation from one sprocket hole to the next within any small group that is 
important. This is one of the reasons for the method of specifying uniformity 
in dimension B. 



Vol 49, No. 2 

American Standard 
Cutting and Perforating Dimensions for 

16-Millimeter Sound Motion Picture 

Negative and Positive Raw Stock 

Reg. U. S. Pat. Off. 

Revision of 




Page I of 2 pages 






0.629 0.001 
0.3000 0.0005 
0.0720 0.0004 
0.0500 0.0004 
0.036 0.002 
30.00 0.03 

15.98 0.03 
7.620 0.0 13 
1.83 0.01 
1.27 0.01 
0.91 0.05 
762.00 0.76 

These dimensions and tolerances apply to the material immediately after 

cutting and perforating. 

*ln any group of four consecutive perforations, the maximum difference of 

pitch shall not exceed 0.001 inch and should be as much smaller as possible. 

(This requirement has been added to the previous standard Z22. 12-1941.) 
fThis dimension and tolerance was given in respect to the center line of the 

perforations in the previous standard Z22. 12-1 941. 
jThis dimension represents the length of any 100 consecutive perforation 



American Standard 
Cutting and Perforating Dimensions for 

16-Millimeter Sound Motion Picture 

Negative and Positive Raw Stock 

Rtg. U. S. Pat. Off. 


Revision of 


Page 2 of 2 pages 


The dimensions given in this standard represent the practice of film manu- 
facturers in that the dimensions and tolerances are for film immediately after 
perforation. The punches and dies themselves are made to tolerances con- 
siderably smaller than those given, but owing to the fact that film is a plastic 
material, the dimensions of the slit and perforated film never agree exactly 
with the dimensions of the punches and dies. Shrinkage of the film, due to 
change in moisture content or loss of residual solvents, invariably results in 
a change in these dimensions during the life of the film. This change is 
generally uniform throughout the roll. 

The uniformity of perforation is one of the most important of the variables 
affecting steadiness of projection. 

Variations in pitch from roll to roll are of little significance compared to 
variations from one sprocket hole to the next. Actually, it is the maximum 
variation from one sprocket hole to the next within any small group that is 
important. This is one of the reasons for the method of specifying uniformity 
in dimension B. 



Vol 49, No. 2 

American Standard 
Cutting and Perforating Dimensions for 

8-Millimeter Motion Picture 

Negative and Positive Raw Stock 

Kr t . v. s. /'/. og. 


Revision of 

Page I of 2 pages 



a a 



a a 



a a 


o- ' *<=> 




a a. 


* , 


a ? o. 





"d" a 


-r- - - - 





0.629 0.001 

15.98 0.03 


0.150 0.0005 

3.810 0.013 


0.072 0.0004 

1.83 0.01 


0.050 0.0004 

1.27 0.01 


0.036 0.002 

0.91 0.05 


Not > 0.001 

Not > 0.025 


0.31 4 0.0015 

7.98 0.04 


0.413 0.001 

1 0.490 0.025 


1 5.000 0.015 

381.00 0.38 




These dimensions and tolerances apply to the material immediately after 
cutting and perforating. 

*ln any group of four consecutive perforations, the maximum difference of 
pitch shall not exceed 0.001 inch and should be as much smaller as possible. 
(This requirement has been added to the previous standard Z22.17-1941.) 

fThis dimension and tolerance was given in respect to the center line of the 
perforations in the previous standard Z22. 17-1941. 

$ln the slitting of double-width film after processing, the cut shall be made 
within 0.002 inch of the center line. 

This dimension represents the length of any 100 consecutive perforation 


American Standard 

Cutting and Perforating Dimensions for * f t.u.s.r. t .og. 

8-Millimeter Motion Picture Revision of 

Negative and Positive Raw Stock Z22.17-1941 

Page 2 of 2 pages 


The dimensions given in this standard represent the practice of film manu- 
facturers in that the dimensions and tolerances are for film immediately after 
perforation. The punches and dies themselves are made to tolerances con- 
siderably smaller than those given, but owing to the fact that film is a plastic 
material, the dimensions of the slit and perforated film never agree exactly 
with the dimensions of the punches and dies. Shrinkage of the film, due to 
change in moisture content or loss of residual solvents, invariably results in 
a change in these dimensions during the life of the film. This change is 
generally uniform throughout the roll. 

The uniformity of perforation is one of the most important of the variables 
affecting steadiness of projection. 

Variations in pitch from roll to roll are of little significance compared to 
variations from one sprocket hole to the next. Actually, it is the maximum 
variation from one sprocket hole to the next within any small group that is 
important. This is one of the reasons for the method of specifying uniformity 
in dimension B. 



Vol 49, No. 2 

American Standard Dimensions for 

16-Tooth 35-Millimeter 

Motion Picture Projector Sprockets 

R'C. V. S. I'al. Of. 

Z22. 35-1947 

Revision of 



Feed Sprocket 


Take-up (Hold Back) 










1.097 0.001 



23.95 0.03 


'.4ol, :o . 

0.943 0.001 




1.097 0.001 
0.932 0.001 


23.67 0.< 


>.401 ; < 


22 Degrees 30 Min 1 .5 Min 

22 Degrees 30 Min 0.75 Min* 

22 Degrees 30 Min 1. 5 M 

Suggested Dimensions 











































*The accumulated error between any 2 teeth not to exceed 4 minutes. 
fThis dimension is the only change from the 1930 edition. 

Aug. 1947 



American Standard 
Cutting and Perforating Dimensions for 

35-Millimeter Motion Picture 

Positive Raw Stock* 

Rrg. V. S. Pat. Off. 

Z22.36- 19.47 

Revision of 

Page I of 2 pages 







D^ T 

* *. 



n * 










1.377 0.001 

34.98 0.03 


0.1 870 0.0005 

4.750 0.01 3 


0.1 100 0.0004 

2.794 0.01 


0.0780 0.0004 

1.98 0.01 


0.079 0.002 

2.01 0.05 


Not > 0.001 

Not > 0.025 


0.999 0.002 

25.37 0.05 


18.70 0.015 

474.98 0.38 




These dimensions and tolerances apply to the material immediately after 

cutting and perforating. 

*This film is used for motion picture prints and sound recording. 

fThis dimension and tolerance was given in respect to the center line of the 

perforations in the previous standard Z22.36-1944. 
iThis dimension represents the length of any 100 consecutive perforation 



American Standard 
Cutting and Perforating Dimensions for 

35-Millimeter Motion Picture 

Positive Raw Stock 

Rrg. U. 5. Pat. Off. 


Page 2 of 2 pages 


The dimensions given in this standard represent the practice of film manu- 
facturers in that the dimensions and tolerances are for film immediately after 
perforation. The punches and dies themselves are made to tolerances con- 
siderably smaller than those given, but owing to the fact that film is a plastic 
material, the dimensions of the slit and perforated film never agree exactly 
with the dimensions of the punches and dies. Shrinkage of the film, due to 
change in moisture content or loss of residual solvents, invariably results in 
a change in these dimensions during the life of the film. This change is 
generally uniform throughout the roll. 

The uniformity of perforation is one of the most important of the variables 
affecting steadiness of projection. 

Variations in pitch from roll to roll are of little significance compared to 
variations from one sprocket hole to the next. Actually, it is the maximum 
variation from one sprocket hole to the next within any small group that is 


featuring a 
Theater Engineering Conference 

and a 
Scientific and Educational Exhibit 

New York, N. Y. 

October 20-24, 1947 

Officers in Charge 

LOREN L. RYDER President 

DONALD E. HYNDMAN Past-President 

EARL I. SPONABLE Executive Vice-President 

JOHN A. MAURER Engineering Vice-President 

M. R. BOYER Financial Vice-President 

C. R. KEITH Editorial Vice-President 

W. C. KUNZMANN Convention Vice-President 

G. T. LORANCE Secretary 

E. A. BERTRAM Treasurer 

General Office, New York 

BOYCE NEMEC . t Executive Secretary 

THOMAS F. Lo GIUDICE Staff Engineer 

MARGARET C. KELLY Office Manager 

HELEN M. STOTE Journal Editor 




James Frank, Jr. 


Gordon A. Chambers, Chairman 

R. T. Van Niman, Vice-Chairman N. L. Simmons, Vice-Chairman 

Herbert Barnett, Vice-Chairman H. S. Walker, Vice-Chairman 


Leonard Satz, Chairman 

Martin F. Bennett, Auditorium Design Charles S. Perkins, Acoustics 
Charles Bachman, Floor Coverings Henry Anderson, Safety and Mainte- 

Donald E. Hyndman, Television nance 

Paul J. Larsen, Television Projection Seymour Seider, Ventilating and Air- 

W. W. Lozier, Lighting , Conditioning; Promotional Display 


Harold Desfor, Chairman 
Leonard Bidwell Don Gillette 


W. C. Kunzmann, Chairman 


Lester B. Isaac, Chairman 

Oscar F. Neu, Chairman 

Lee E. Jones, Chairman 


H. F. Heidegger, Chairman , 

Assisted by Members of the New York Projectionists Local 306 


M. W. Palmer, Chairman 


Harry B. Braun, Chairman 

Robert T. Kenworthy, Exhibit Manager 

W. W. Simons, Chairman 
Sidney B. Moss J. W. Servies 

Aug. 1947 




Monday, October 20, 1947 

9 : 00 A.M.-10 : 00 P.M. PENN TOP 

Scientific and Educational Exhibit. 
Registration. Advance sale of 

Luncheon and Banquet Tickets. 

Business Session. 

Theater Engineering Session, "Intro- 


Get -Together Luncheon (Eminent 


Theater Engineering Session, "Phys- 
ical Construction". (Regular, 
Prefabricated, and Drive-in 

Theater Engineering Session, "Audi- 
torium Design". 

Tuesday, October 21, 1947 

9 : 00 A.M.-10 : 00 P.M. PENN TOP 

Scientific and Educational Exhibit. 
9:00 A.M. HOTEL, 18TH FLOOR 
Registration, Advance Sale of Ban- 
quet Tickets. 

General Technical Session. 

Wednesday, October 22, 1947 

Theater Engineering Session, "Floor 


Theater Engineering Session, "Tele- 

9 : 00 A.M.-5 : 00 P.M. PENN TOP 

Scientific and Educational Exhibit. 
9:00 A.M. HOTEL, 18TH FLOOR t 
Registration, Advance Sale of Ban- 
quet Tickets. 

General Technical Session. 
Theater Engineering Session, "Light- 

7: 15 P.M. GEORGIAN ROOM (Recep- 
tion Foyer) 

Cocktail Hour for Holders of Ban- 
quet Tickets. 

62nd Semiannual Banquet and eve- 
ning for social get-together (danc- 
ing and entertainment). Tables 
may be reserved at the Registra- 
tion Headquarters prior to noon of 

October 22. 

Thursday, October 23, 1947 
9 : 00 A.M.-10 : 00 P.M. PENN TOP Theater Engineering Session, "Acous- 

Scientific and Educational Exhibit. tics". 


2:00 P.M. SALLE MODERNE Theater Engineering Session, "Tele- 

Friday, October 24, 1947 

9 : 00 A.M.-5 : 00 P.M. PENN TOP 

Scientific and Educational Exhibit. 
General Technical Session. 


Theater Engineering Session, "Safety 

and Maintenance". 

Theater Engineering Session, "Ven- 
tilating and Air-Conditioning; 
Promotional Display". 



Hotel Reservations and Rates 

The management of the Hotel Pennsylvania, 33rd Street and Seventh Avenue, 
Convention Headquarters, extends to members and guests of the Society of 
Motion Picture Engineers the following per diem room rates, European plan : 
Room with bath, 1 person: $4.00, $4.50, $5.00, $5.50, $6.00, $6.50, $7.00. 
Room with bath, 2 persons, double bed: $6.00, $6.50, $7.00, $7.50, $8.00, $8.50, 

Room with bath, 2 persons, twin beds: $7.00, $8.00, $8.50, $9.00, $10.00, 

$11.00, $12.00. 
Parlor Suite, for 1 or 2 persons: $13.50, $14.50, $16.50. 

NOTE : Room accommodations must be booked early and direct with Frank 
A. Morse, front office manager, Hotel Pennsylvania, and prior to October 15. No 
rooms will be assured or available unless confirmed by the hotel management. 


The Conference Registration Headquarters will be located on the 18th floor 
of the hotel adjacent to the Salle Moderne, where all business and technical 
sessions will be held during the five-day conference. Members and guests must 
register to attend sessions and the Exhibit. The fee is used to defray Conference 


Members and guests who contemplate attending the 62nd Semiannual Conven- 
tion and Theater Engineering Conference in New York should consult their local 
railroad, Pullman, and airline agents at least 30 days in advance of departure date 
regarding effective schedules and rates. 

SMPE Get-Together Luncheon 

The usual Get-Together Luncheon will be held in the Georgian Room, on 
Monday, October 20, 1947, at 12:30 P.M. 

The luncheon program and eminent guest speakers will be announced later. 
Guaranteed seating at the luncheon will be assured only if tickets are procured 
prior to October 17, 1947. Assist the Committee and hotel in providing ac- 
commodations by complying with this request. 

Informal Banquet and Dance 

The SMPE 62nd Semiannual Informal Banquet and Social Get-Together will 
be held in the Georgian Room of the Hotel Pennsylvania on Wednesday evening, 
October 22, at 8:30 P.M. (Dress optional.) 

A Cocktail Hour for holders of Banquet tickets will be held in the Georgian 
Room (Reception Foyer) preceding the Banquet from 7:15 to 8:15 P.M. 

Banquet tickets should be procured and tables reserved at the Registration 
Headquarters prior to noon on October 22. The Banquet program will be an- 
nounced later. 

Luncheon and Banquet tickets may be procured in advance of the dates 
of these functions through the Society office or through W. C. Kunzmann, Con- 
vention Vice-President, who will be at the Pennsylvania a week prior to the open- 
ing date. 

NOTE: All checks or money orders issued for registration fee and luncheon or 
banquet tickets should be made payable to W. C. Kunzmann, Convention Vice- 
President, and not to the Society. 



The editors present for convenient reference a list of articles dealing with subjects 
cognate to motion picture engineering published in a number of selected journals, 
Photostatic or microfilm copies of articles in magazines that are available may be 
obtained from The Library of Congress, Washington, D. C., or from the New York 
Public Library, New York, N. Y., at prevailing rates. 

American Cinematographer International Projectionist 

28, 6 (June 1947) 22, 6 (June 1947) 

Magnetic Tape Recording (p. 199) Acetate Stock to Supplant Nitrate 

Painting with Technicolor Light (p. 5) H. B. SELLWOOD 

(p. 200) H. A. LIGHTMAN Television Today and Its Problems 

The Cinema Workshop. Special (p. 8) A. N. GOLDSMITH 

Effects (p. 208) C. LORING 

A New Series of Camera Lenses for Tele-Tech 

16-Mm Cinematography (p. 210) 5^ Q (June 1947) 

W. B. RAYTON Color Television for Theatres (p. 44) 

International Photographer K ' G \ SHEA 

19, 6 (June 1947) Embossing Type Sound Recorder 

The Baltar Series of Lenses (p. 12) <P; 55) 

A. E. MURRAY Bibliography of Disc Recording (p. 

Zoomar Lens Makes Debut (p. 22) 73) A ' J ORYSZ 


Radio News 

38, 1 (July 1947) 

The Recording and Reproduction of 
Sound. Pt. 5 (p. 55) O. READ 


The JOURNAL of the SMPE wishes to correct the following errors which 

appeared recently in its pages. 

* * * 

At the bottom of page 311, in the April, 1947, issue, the following statement was 
made: "Finally, at Beaconsfield will be housed the Central Film Library." 
The British Information Services has brought to the attention of the Editor that 
the address of the Central Film Library continues to be: Imperial Institute, 
South Kensington, London, S. W.7, England. 

* * * 

On page 418 of the May, 1947, issue of the JOURNAL, alongside of the heading 
"July 1928", there appears this statement. "Paramount began recording in 
Hollywood on a temporary channel and first used sound in their picture 'Warming 
Up', with Richard Dix." In a letter to the Editor, Mr. Walter F. Wanger states, 
"I was in charge of production at Paramount at that time and we did not begin 
recording in Hollywood but recorded on disks at the Victor Talking Machine 
Plant in Camden, N. J." 


Oiveral more American Standards on Motion Pictures, 
printed in the new 8!/2- x 11 -inch format, will soon be available for inclusion 
in the SMPE Standards binder, shown above. These Standards, some of 
which have been published in recent issues of the Journal, are supplied as 
a service to motion picture engineers, industrial librarians, and to those in 
the industry wro need to maintain up-to-date files of American Motion 
Picture Standards for easy and ready reference. 

The price of the SMPE Standards Binders, with a complete set of 32 
Arrrerican Motion Picture Standards, is only $6.10.* As a further service, 
purchasers of these Binders are notified by the Society when new Standards 
or revisions thereof are published. All American Motion Picture Standards 
published in the future will be punched to fit this binder. 

Send check, money order, or company purchase order now to the 
Society of Motion Picture Engineers, Hotel Pennsylvania, New York 1, N. Y. 

* Add 50 cents for postage and special packing if mailed outside of fhe United States. If mailed to 
New York City address, add 2 per cent Sales Tax. 


Vol 49 SEPTEMBER 1947 No. 3 


Retooling for Education 1948 W. A. WITTICH 187 

Educational Films for a Democratic Tomorrow 


Psychology of the Sound I^ilm L. MERCER FRANCISCO 195 
Training-Film Production Problems REID H. RAY 203 

Light Generation by the High-Intensity Carbon Arc 


Motion Picture Screen Light as a Function of Carbon- 
Arc-Crater Brightness Distribution M. T. JONES 218 

Adaptations and Applications of 16-Mm Motion Pic- 
ture Equipment to Medical and Scientific Needs 

Kodachrome Motion Pictures of the Human Air and 
Food Passages 


Sound Absorption and Impedance of Acoustical 
Materials HALE J. SABINE 262 

Report of the Studio Lighting Committee 279 

Committees of the Society 289 

Current Literature 296 

Society Announcement 296 

Copyrighted, 1947, by the Society of Motion Picture Engineers, Inc. Permission to republish 
material from the JOURNAL must be obtained in writing from the General Offide of the Society. 
The Society is not responsible for statements of authors or contributors. 

Indexes to the semiannual volumes of the JOURNAL are published in the June and December 
issues. The contents are also indexed in the Industrial Arts Index available in public libraries. 










** President: LOREN L. RYDER, 

5451 Marathon St., Hollywood 38. 
** Past-President: DONALD E. HYNDMAN, 

342 Madison Ave., New York 17. 
** Executive Vice-President: EARL I. SPONABLE, 

460 West 54th St., New York 19. 
^Engineering Vice-President: JOHN A. MAURER, 

37-01 31st St., Long Island Cky 1, N. Y. 
** Editorial Vice-President: CLYDE R. KEITH, 

233 Broadway, New York 7. 
^Financial Vice-President: M. RICHARD BOYER, 

E. I. du Pont de Nemours & Co., Parlin, N. J. 
**Convention Vice-P resident: WILLIAM C. KUNZMANN, 

Box 6087, Cleveland 1, Ohio. 
** Secretary: G. T. LORANCE, 

63 Bedford Rd., Pleasantville, N. Y. 
* Treasurer: E. A. BERTRAM, 
850 Tenth Ave., New York 19. 

**JOHN W. BOYLE, 1207 N. Mansfield Ave., Hollywood 38. 

*FRANK E. CARLSON, Nela Park, Cleveland 12, Ohio. 

*ALAN W. COOK, Binghamton, N. Y. 
**ROBERT M. CORBIN, 343 State St., Rochester 4, N. Y. 
**CHARLES R. DAILY, 5451 Marathon St., Hollywood 38. 
"fjAMES FRANK, JR., 356 West 44th St., New York 18. 

"JOHN G. FRAYNE, 6601 Romaine St., Hollywood 38. 
**DAVID B. JOY, 30 East 42d St., New York 17. 

*PAUL J. LARSEN, 1401 Sheridan St., Washington 11, D. C. 

*WESLEY C. MILLER, MGM, Culver City, Calif. 
**HOLLIS W. MOYSE, 6656 Santa Monica Blvd., Hollywood. 
*JA. SHAPIRO, 2835 N. Western Ave., Chicago 18, 111. 
*WALLACE V. WOLFE, 1016 N. Sycamore St., Hollywood. 

*Term expires December 31, 1947. tChairman, Atlantic Coast Section. 
**Term expires December 31, 1948. tChairman, Midwest Section. 
"Chairman, Pacific Coast Section. 

Subscription to nonmembers, $10.00 per annum; to members, $6.25 per annum, included in 

their annual membership dues; single copies, $1.25. Order from the Society at address above. 

A discount of ten per cent is allowed to accredited agencies on orders for subscriptions 

and single copies. 
Published monthly at Easton, Pa., by the Society of Motion Picture Engineers, Inc. 

Publication Office, 20th & Northampton Sts., Easton, Pa. 

General and Editorial Office, Hotel Pennsylvania, New York 1, N. Y. 

Entered as second-class matter January 15, 1930, at the Post Office at Easton, Pa. 

under the Act of March 3, 1879. 


Vol 49 SEPTEMBER 1947 No. 3 


Summary. The problem of educational responsibility is outlined, particularly 
the aspects concerned in the enormous increase in scope of man's environment during 
the past hundred years. The means for becoming acquainted with this expanded 
environment are discussed and the place of audio-visual teaching aids in this broad 
over-all picture is examined. 

It has been said that the primary responsibility for those who 
assume assignments in our classrooms and on our administrative staffs 
is to organize the experiences of our environment so as to allow young 
learners to become acquainted with those things which lie about them 
and among which they live as methodically, effectively, and as realis- 
tically as possible. Thus our primary responsibility for education is 
to acquaint young learners with their environment to lead the chil- 
dren who come under our direction through a series of experiences 
which will allow them to become informed completely about two 
aspects of the world in which they live : First, those things and ex- 
periences which are included in the environment which nature has 
provided ; and second, those plans, those cultural patterns, and those 
problems which go to make up the man-made or social environment. 

This may seem an oversimplification, but when we analyze it, all 
of man's relationships are in terms of his ability first to understand 
those things which exist within these two environmental spheres. 
On the basis of his complete understanding of these, he attempts to 
arrive at a course of action which will allow him to experience happi- 
ness, peaceful cultural relationships, and a high standard of living. 
In contrast to this, we as adults have periodically allowed ourselves 
to be sent crashing into the ruins of great world conflicts. This ex- 
perience cannot and must not reoccur tomorrow. 

* Presented Apr. 23, 1947, at the SMPE Convention in Chicago. 
** Director, Bureau of Visual Instruction, University of Wisconsin, Madison. 


188 WlTTlCH Vol 49, No. 3 

Today those of us who are searching for means of acquainting our 
young people with formal educational experiences are confronted with 
problems which did not exist a century ago, but which today are being 
placed squarely in the laps of educators. Today education is the sole 
remaining social agency capable of assuming this responsibility. 
That education is today being the recipient of this heretofore un- 
realized social responsibility is evidenced on many fronts. 

First, in the report of the United States Chamber of Commerce we 
have an outright acknowledgment that in those sections of the coun- 
try where good educational systems exist, there, too, exists a high 
standard of living. Second, from this same source we have the 
acknowledgment that in those communities where a good system of 
public education exists, there, too, exists a profitable market for the 
commodities which man produces. Third, for the first time in our 
history has the president of the United States formally announced 
through his radio talks that upon the assumption by educational 
agencies of the responsibility for world enlightenment lies a last re- 
maining hope for peace and for a world order in which we can seek to 
attain desirable social outcomes. 

These three evidences from sources, heretofore known to be some- 
times antagonistic and certainly disinterested toward the place which 
formal education occupies, are evidences of a growing popular trend, 
and a change in attitude toward public education. Why has this 

Allow us to see how the role of education has changed over the last 
hundred-year period. One hundred years ago what was the "com- 
munity" and what was the responsibility of education to that com- 
munity? Many of us come from southern states in which the county 
government division was formed on the basis of the distance a man 
could ride on horseback during a one-day period. In the north, where 
hilly terrain predominated, the local unit of government of the town- 
ship was frequently determined by the distance a man could see the 
rim of the valley or the land bounded by rivers or hillcrests. Within 
these limited communities, the school accepted its responsibility its 
responsibility for the teaching of arithmetic to the rule of three, of 
writing, and of reading and as soon as the child could be turned over 
to an apprentice situation, he was there to learn for himself, an the 
shortest possible time, how to earn a living within the community in 
which he was reared. That was education a hundred years ago. 
That was the community that education served. 

Sept. 1947 RETOOLING FOR EDUCATION 1948 189 

Without going any farther to trace the trend from that time to the 
present, examine briefly what the problems of an enlarged world- wide 
environment mean to the responsibility which education must 
accept today. Today when we speak in terms of United Nations, 
when we speak in terms of girdling the globe as Reynolds did in 72 
hours, when we speak in terms of rocket missiles which will accom- 
plish the transit of the oceans in a matter of minutes, when we speak 
of communication systems undreamed of a decade ago, we know that 
the responsibility for man understanding his environment is no longer 
confined to the valley, to the county, or to the township, but rather 
to the whole world. Today's educational responsibility encompasses 
a global environment. Today's school responsibility is beyond all 
previous comprehension. It is so vast that today's society is gladly 
willing that public education assume a role which it has never been 
asked to assume in the past. Thus, we in education find ourselves 
members of the last remaining social agency capable of approaching 
this tremendous problem. 

Let us stop to examine the tools with which we have to work, the 
personnel which staffs our school, and ask ourselves, "Are we up to 
the problem?" Very largely, our teaching staffs comprise those per- 
sons who were with us ten years ago. Very largely, our curriculum 
plans are unchanged. Unchanged, too, is the instructional equip- 
ment and the educational tools with which we have to work. Our 
responsibilities have leaped ahead in inaccountable terms, but the 
tools with which we work are dragging pitifully. In the words of 
Ernest Horn of Iowa, "90 per cent of our classroom learning experi- 
ences are confined to a textbook." In the field of geography we are 
still clinging to a course of study which was developed twenty years 
ago and which does not take into consideration an understanding of 
parts of the globe that have recently leaped into prominence during 
the last five-year period. The problem remains, what are we going to 
do? What tools can we locate? 

Ten years ago, even today, the philosophic argument still remains, 
"Should teachers be satisfied to pass along to young learners an his- 
torical accumulation of information out of the past, or should teachers 
be allowed to teach of the present and plan for the future?" I be- 
lieve that the last is our role. 

Today this country is the single remaining first-line political and 
economic power. This country alone must assume full responsibility 
for the rehabilitation of the entire world. The school certainly must 

192 ALBERT Vol 49, No. 3 

misunderstood. Probably the best division of the film medium is 
into the two categories of theatrical (for showing in theaters) and 
nontheatrical (for showing in schools, churches, civic organizations, 
and union halls) . Theatrical releases are designed primarily for en- 
tertainment and are not our concern here. Our concern is with the 
non theatrical field, with governmental, industrial, and sponsored or 
unsponsored classroom films. This is a complex, vital new means of 
communication. In prewar America there were two decades during 
which we felt our way in the field. In general, films were dull and 
technically poor. There were few projectors for nontheatrical show- 
ings. School budgets did not include funds for visual tools and 
equipment. With a limited and unpredictable market, a dozen repu- 
table firms made low-cost films for schools, and a dozen dozen industries 
advertised themselves and their products by objectionably sponsored 
films. But it was not an established industry. Distribution outlets 
were wholly inadequate. It was a formative period, with isolated 
groups toying with an idea whose potentialities were neither recog- 
nized nor exploited. 

The war brought films to a position of significance because they 
were the only logical way to speed up the training of our vast numbers 
of army and navy personnel-, so suddenly taken from civilian life and 
fashioned into military men. All governmental and civilian agencies 
united to produce a flood of training and propaganda films. Thus an 
emergency brought proof to what had before been supposition; the 
film medium came into its own as a teaching weapon. 

The facts about use of films in education are well known. Out of 
the war, in addition to this knowledge, came a multitude of men and 
women trained in the field, sold on its future, and eager to make films 
their postwar vocation. And what do they find? 

We as a nation of civilians are not ready either to make or to use 
films in the manner of a nation geared for war. Those of us in the 
nDntheatrical field entered the era after the war full of hope ; now, more 
than two years later, certain truths stand out. The school market 
has remained a fairly static thing. With a few heartening exceptions, 
budgets are low. There are only a few thousand projectors in use in 
American schools. The wealth of talent trained in war has formed 
literally hundreds of production companies, most of them operating 
with limited capital, and they have found (a) that an investment in 
an educational film is a long-range and uncertain thing, and (b) 
that industrial and governmental sponsors are difficult to sell. 


Distribution, the largest single handicap, is uncharted and inadequate. 
Poor distribution means a low print sale ; a low print sale means a re- 
duced production budget ; a reduced production budget means tech- 
nical excellence far below that of theatrical releases. And so the cycle 
continues, with schools and other consumers waiting for a better prod- 
uct, and the producer unable to make a better product until he 
has a larger potential market. 

These considerations are not in any sense an indictment of the field 
or its potentialities, nor are they a reflection of waning enthusiasm on 
the part of our company. On the contrary, .we are more positive 
than ever that the nontheatrical film will play a significant role in 
shaping the postwar world. . 

There are corresponding encouraging things. Across the country 
educators are meeting and working to bring use of visual aids to a 
prominent position in schools. Manufacturers are perfecting low- 
cost, lightweight, easy-to-operate projectors. Foundations and 
industries are sponsoring more and more films in the public interest. 
Scientific research, such as that conducted under the auspices of the 
Motion Picture Association, is determining what kinds of film are 
needed and what kinds of film teach best. New techniques are en- 
larging the scope of the motion picture. Labor is beginning to use 
film as a means of expressing its point of view. The church, too, is 
making use of films to help retain its position in a changing world. 
The Department of State, the Department of Agriculture, the 
Attorney-General these and many more governmental agencies are 
making films for national and foreign distribution. New fields are 
being opened social hygiene, medicine, psychology. These things 
make us know that the job is worth doing, and worth doing well. 

Ours is an important responsibility- It is obvious to even a casual 
observer of fourth-decade twentieth-century living that much must be 
done, and soon, if the democratic way of life is to survive. The 
danger to our country lies not so much in the encroachment of various 
"isms" and ideologies as it does in our failure, individually and col- 
lectively, to understand and participate actively in our government. 
We have fallen into the terrifying complacency of letting others think 
for us. This is because we are not educated to our responsibilities. 
America began with the idea of giving to every man an equal chance, 
founded on a belief that the majority of common men, properly in- 
formed, will choose and enforce the right kind of governmental action. 
Thomas Jefferson stated it this way : 


"I know of no safe depository of the ultimate powers of society 
but the people themselves ; and if we think them not enlightened 
enough to exercise their control with a wholesome discretion, the 
remedy is not to take it from them, but to increase their discretion 
by education." 

The need, then, is for mass education. Education is largely a matter 
of communication. Film is best suited for this communication of 
ideas because it best combines four essentials : it is simple, it is fast, 
it is universal, it is democratically sound. As George Bernard Shaw 
wrote, "The number pf people who can read is small, the number of 
those who can read to any purpose is smaller, and the number of those 
too tired to read after a hard day's work is enormous . . . but all except 
the blind and the deaf can see and hear." 

What kind of films ? How can it be done ? What is the next move ? 
The answer lies in our ability to agree on one simple premise. I be- 
lieve that it is vital to educate Americans so that they will participate 
actively and intelligently in a democratic government. If you agree 
with me, we have a starting point. I recognize the importance of 
using films to teach fundamentals of arithmetic and science ana! social 
studies, to advertise our products, to train our personnel, to build good 
will. These things are all significant. But we must place the empha- 
sis on films which ... on all levels, to all peoples, in all areas . . . teach 
citizenship the kind of citizenship which results in the average 
American's being an informed, tolerant, active participant in the 
affairs of his country and his world. 

Here are some specifics, film subjects which should be converted 
into motion pictures for America to see and hear: Housing: Juve- 
nile Delinquency. Healthy Daily Living. Child Training. How 
Your Vote Is Reflected in Laws. Family Relationships. Manage- 
ment and Labor. Household Management. Marital Hygiene. 
Government in Local and World Affairs. Racial Equality. Inter- 
national Relations. There are a hundred such subjects and they will 
not be made into films soon enough, or well enough, unless the urgency 
of our crisis is brought home to those individuals, groups, industries, 
foundations, and governmental agencies which can invest America's 
money in America's future. 

We sincerely believe that unless we in the motion picture industry 
combine our efforts with those of others working to educate America, 
we may find that there will be no democracy tomorrow, and, for some 
of us, no tomorrow at all. 


Summary. A s an educational instrument the sound film is particularly effective 
for influencing people in groups, bringing to bear factors in social as well as indi- 
vidual psychology: social facilitation, the impression of universality, and prestige. 
The physical conditions under which its message is received and the absorption of the 
audience in the continuity of the screen story polarize the attention almost to the degree 
of hypnosis, with corresponding effect upon the subconscious and memory. Little in- 
tellectual effort is required to comprehend the meaning of visual-action images when co- 
ordinated with spoken words. The appeal to the emotions is equally strong through 
the control of empathic responses by means of meaningful real-life situations dra- 
matically portrayed. Self-identification by the audience with the screen characters 
and subject matter provides vicarious experience, short-cuts the learning process, and 
arouses the will to believe or to act. 

Signs of the times indicate that the engineer is beginning to show 
some concern over the social implications of his inventions and to feel 
some responsibility for them. 

Dr. Lee de Forest recently registered a vigorous protest against the 
abuses which his great invention, radio, has suffered at the hands of 
its commercial exploiters. 

The Association of Atomic Scientists has made it plain that they 
do not feel that their responsibility for the atomic bomb ended with 
their invention of it. They have become quite vocal in their in- 
sistence that atomic energy be applied to social advancement rather 
than to the destruction of civilization. 

The first note made of the misuse of the engineer's talents was re- 
corded by Thorstein Veblen twenty-six years ago in a trenchant little 
book, The Engineers and the Price System. Few engineers seemed to 
have paid any attention to it, however, until 1946, when John Mills, 
of the Bell Telephone Laboratories, published The Engineer in Society. 
How many engineers have read Mills' provocative book remains to 
be seen. 


The motion picture engineer has apparently shown no more con- 
cern over the uses to which his products have been put than other 
technologists, in spite of the fact that the motion picture has been 

* Presented Apr. 23, 1947, at the SMPE Convention in Chicago. 
** Francisco Films, Chicago, 111. 


196 FRANCISCO Vol49, No. 3 

more severely criticized for the ill effects it has had upon our culture 
and mores than any other invention. Whole volumes of criticism 
have been published and research foundations have devoted their 
energies to studying the baneful influence of motion pictures upon 
attitudes, social life, and behavior patterns. 

In view of the fact that the motion picture, particularly the sound 
film, ranks with the invention of movable type, the telephone, the 
radio and the automobile, as an instrument of communication which 
has had far-reaching effects upon our way of life, it behooves the tech- 
nologist working with the medium to extend the range of his interest 
in it beyond the confines of the engineering laboratory into the realm 
of the uses of the film. 


To date, the motion picture has been used almost entirely as an 
entertainment device, for that is where the "big money" has lain. 
Within the past decade, however, the application of the filmic me- 
dium to other ends of communication education, instruction, and 
propaganda, in its positive sense indicates that the future useful- 
ness of the medium in this field may transcend in importance its appli- 
cation to entertainment. The social scientist, particularly the social 
psychologist, has discovered unequaled powers in the sound film to 
influence the minds and emotions of people for better or for worse. 


The producers of sound films for commercial and educational pur- 
poses are probably better acquainted with the psychological impact 
of the sound film and the techniques for producing desired attitudes 
and responses with it, than any other factors in the motion picture 
industry. In the creation of sound films for educational, sales, or 
training purposes, they begin with a study of the audience, the occa- 
sion, the objective of the film, and the idea content for it, and they 
employ the various elements of the medium to create the desired im- 
pression upon the mind and nervous system of the audience under the 
given conditions of the screening and of the audience's mental "set" 
toward the film's subject matter. 

To appreciate fully the significance of this approach, try designing 
a sound film. It will be quickly realized that in the sound motion 


picture something more is involved than cameras, film stock, and 
sound channels. Probably the most important lesson that the Armed 
Services learned in their vast film program was that the creation and 
production of sound films had best be left to those who best under- 
stand the medium as an instrument of expression ; therefore as a tool 
for the shaping and control of psychological factors. George Arliss, 
whose experience on the stage, in silent motion pictures, in sound 
movies, and in writing for publication, qualified him to speak, vividly 
characterized the uniqueness of the sound film as an art form. He 
said that there is a greater difference between the sound motion pic- 
ture and the silent motion picture, as a medium of expression, than 
there is between the silent motion picture and the printed page. He 
averred that he had to learn to act all over again when sound came 
into pictures. 


The sound film is, by its very nature, a medium for influencing 
minds in groups, whereas other mediums of communication, such as 
the printed page, posters, and radio, influence minds as individuals. 
The reader of a magazine or the listener to a radio program reacts to 
the message that reaches his mind as an individual, whereas the 
member of a sound-film audience reacts to the film as a unit of a co- 
acting group. Social psychology, as well as individual psychology is, 
therefore, at work in the film audience. 


Consider, for a moment, the psychology of the audience situation of 
the sound film. The various members of the audience are seated in 
rows, all facing toward a common source of sensory impressions. 
This regimentation of their bodies tends to induce a corresponding 
regimentation of their minds. In the regimented film audience, 
"social facilitation" is at work; that is to say, the reactions of one 
member of the group are "sensed" by his neighbors and tend to spread 
from one to the other, like a contagion. A related phenomenon re- 
sults the "impression of universality", or the conviction that what 
is true for one individual is true for another. This is a vital factor in 
credibility. Of course, social facilitation and the impression of uni- 
versality prevail in almost all audiences, but not to the degree that 
they do in the sound-film audience, because of other factors in the 
medium which will be discussed later. 

198 FRANCISCO Vol 49, No. 3 


The sound film, as a medium of communication, has prestige, an 
important quality in any medium. Its prestige is not derived solely, 
however, from its association with Hollywood and the glamour of 
sound-film-production technology. Most of its prestige is derived 
from the unique qualities inherent in the medium itself, particularly 
the intensely concentrated attention it demands, which will be re- 
ferred to later. Prestige is also derived from the occasion of the film 
showing : the audience has assembled especially to see and hear the 
film and everybody submits to its dominating influence. The au- 
thority of the spoken word and the vivid visual-action images that are 
burned into the mind, add to the prestige of the medium. There is 
no other medium of expression the largest, most popular magazine, 
the world's largest spectacular, or the biggest national network that 
can match the impact of the sound film in so far as the prestige of the 
medium is concerned. 


There are psychological factors in the conditions under which the 
sound-film message is received that are no less important than those 
of the audience situation. Consider the fact of the darkened room, 
with all extraneous sounds shut off, in which the sound film does its 
work. The two primary sense organs those of sight and hearing 
are focused on a single source of sensations : the screen and the loud- 
speaker. The effect created is akin to that of hypnosis, for, as you 
know, to put a person "out", the hypnotist induces him to focus the 
attention of his eyes on a single point while he drones something into 
the subject's ears. The sound film, likewise, "polarizes" the atten- 
tion to almost the same degree and induces in the subject a state of 
suggestibility similar to that induced by the hypnotist. 

Scientific proof of the intensity of the concentration of the attention 
of the sound-film audience to the influence of the screen and loud- 
speaker has been provided on many occasions, by many experi- 
menters. They have flashed bright lights and rung bells to distract 
the attention of the sound-film audience and have taken pictures of 
the audience's reactions. These have shown nothing but momentary 
glances away from the screen. The mind that "wanders" along any 
paths except those directed by the screen story is alniost unknown in 
the sound-film audience. The salesman or propagandist who wishes 


to hold attention to his sales message can do it beyond peradventure 
in this dynamic medium. ''Spellbound" is not too strong a word for 
describing the attention-compelling power of the sound film. 


The sound film sustains attention for other reasons than the 
physical conditions of the screening in the darkened room with extra- 
neous sounds shut out. The medium itself grips the attention through 
the "flow" of the ideas it presents. It presents visual-action images 
in a "stream-of -consciousness" manner, requiring virtually no intellec- 
tual effort for comprehension. If the presentation is in the form of a 
drama unfolding in the words and actions of screen characters with 
whom the audience can identify themselves readily, then, indeed, the 
attention is spellbound, for the audience becomes "lost in the story". 
Psychologists say that the mind cannot concentrate on any fixed 
object longer than about seven seconds. In the sound motion picture 
there is no fixed object; even in the sound slide film, if it is properly 
conceived, there is a change of picture on the screen about every seven 
seconds, so the eye simply cannot stray without missing something, 
and the mind cannot wander. 


One might, conceivably, be entranced by a continuously moving 
object and still not understand it. What can be said, therefore, as to 
the impact of the sound film upon the mind? This: that it makes 
its idea content crystal clear, because it presents ideas in visual-action 
images of the very type which is believed to be involved in thought 
processes themselves. The Chinese made this point thousands of 
years ago and expressed it in what has become a cliche" in the field of 
the graphic arts in "one picture is worth ten thousand words". The 
great significance of this point in the educational field has recently 
been brought out in the work of the semanticists. They have dis- 
covered that "verbalism" is the most serious defect of our educational 
methods and they see in the sound film a means of correcting it. 


The importance of words should not be underrated, for words, 
rather than pictures, are the symbols of thought. A film without 
words is relatively meaningless. In every effective sound film the 

200 FRANCISCO Vol 49, No. 3 

idea content is presented primarily in words and secondarily in pic- 
tures; the pictures supplement, complement, define, and clarify the 
meaning of the words. But the words of the sound film are spoken 
words. Spoken language is the kind that all of us use every day, all 
day long, and that all of us hear all the time. They are readily com- 
prehended, assuming they are within the range of our listening if not 
of our speaking or writing vocabulary. In fusing these two primary 
factors in clarity visual-action images and spoken words the 
sound film reaches the intellect more readily, with less effort, and 
more impressively than does any other method of expression. 


People do not, however, think with their minds alone. The be- 
haviorist insists that they also think with their viscera and the neu- 
rologist that their endocrine glands are involved. Everyone will 
admit that people feel more than they think and that the appeal to the 
emotions is often of more importance than the appeal to the mind. 
The impact of the sound film on the emotions is, if possible, even 
greater than its impact on the mind. Attitude, or ' 'mental set' ' , is the 
determinant of mental activity, and it is the product of the emotions. 
The recent rapid strides that have been made in scientifically pretest- 
ing the appeal of sound films is based upon measurements of emotional 
responses, by instruments something similar- to the lie detector. 

Probably the most important element in the sound film influencing 
the emotions is its sound. The effect of sound upon the feelings is 
most readily appreciated in the field of music . ' 'Music hath charms to 
soothe the savage breast." That sound affects the feelings or emo- 
tions, while sight appeals to the mind, has always been an accepted 
fact. The birth of Christ was announced to the Wise Men of the East 
men of intellect through a star in the heavens, a visual symbol, 
whereas to the shepherds feeding their flocks presumably simple- 
minded folk it was heralded through angels' voices singing. When 
the history of music is written a generation from now, the contribu- 
tions which the sound film has made to it will be better appreciated 
than it is today. Every script writer, even of commercial films, 
knows what music can do to advance his story and to induce the de- 
sired attitude in the film audience. Along with music, of course, is in- 
cluded "sound effects" ; they help the sound film, as one writer puts it, 
to "create the fury as well as the battle, the song as well as the lark". 



The impact of the sound film upon the emotions is manifest in 
other elements, however, as well as in its music and sound effects. 
The action of screen characters involved in tense, emotional situa- 
tions, induce corresponding emotions in the audience. 

The psychologists have a word for this phenomenon of inducing 
emotions in one individual through portraying emotions in another. 
It is "empathy". It is your empathic responses which make you al- 
most push your neighbor off his seat at a football game, when your 
own body leans rigidly in the direction of the line plunge of your own 
team. It is empathy that makes your muscle tensions follow those of 
the screen character in the dramatic portrayal. 


So strong is the appeal of the sound film, through both its sound and 
its visual-action images, to the emotions, that it often is the equal of 
real-life experience itself in intensity. Indeed, there are films in 
which the screen-presented story seems even more real than real life ! 
Imagine what that means in propaganda, selling, public relations, 
education, and training. It means that, with the sound film, you can 
groove or condition the nervous systems of people, in a directed, con- 
trolled manner, almost as well as experience itself! You can provide 
them with vicarious experience in the form of muscle tensions, nerv- 
ous responses, blood pressure, respiration, and all the acitivities of the 
sympathetic nervous system by controlling the secretions of their 
endocrine glands through the stimulus of the sound film. What other 
medium of communication can even approach the sound film in this 


Earlier in this article reference was made to the "polarization" of 
attention which the sound film induces, by way of the darkened room 
with extraneous sounds shut out and the gripping hold of the screen 
story. This phenomenon has profound effects upon the memory. 
Psychologists tell us that there can be no learning without memory. 
Sales or educational material that is presented in the sound film pene- 
trates into the depths of the subconscious mind of the individual who 
is "lost" in its story, and every item of the material is surrounded by 
a rich background of associated materials which serve to aid recall 


long after the sound-film showing. Having provided both visual 
images and aural impressions with the material, recognition of the re- 
called elements is instant and easy. 

Proof of the lasting effects of the sound film upon the memory has 
been provided in abundance by many kinds of tests, some of them 
revealing facts hard to explain. For example, in some instances, ma- 
terial from a sound film is more readily recalled several months after 
exposure to it than it is twenty -four hours or even immediately after- 


Advertisers who have been told that anything said on the radio has 
to be repeated at least three times in order to make it sink in, are go- 
ing to have to learn restraint in their commercials when they get into 
television or when they use sound motion pictures unless to irritate 
is their purpose. They are going to find that the less often they say 
what they have to say about their product in the sound film, the more 
favorable impression they will create ; indeed, they may not have to 
say anything, if the pictures they employ say it without words. They 
will also have to learn that the screen audience abhors repeated pic- 
tures. It responds against subject matter with even more intensity 
than it responds in favor of it, a fact that some sound film producers 
know but that many users have yet to learn. 

It can be appreciated readily how markedly the sound film differs, 
in its psychological factors, from other mediums of expression, the 
printed word, the radio, the lecture, and the stage. Students of social 
psychology, of educational problems, of propaganda, of public rela- 
tions, and of training procedures, are fast recognizing the sound film 
as the most potent instrument for accomplishing their objectives that 
has been developed in the field of communication. 


Summary. Tke production of motion pictures for training presents distinct 
reasons for- techniques different from ordinary documentary or factual film methods. 
This is true whether the visual aid is produced during the stress of wartime require- 
ments, or for training in industry, or educational fields in normal times. There are 
problems in (1) set construction, (2) camera angles, (5) selection and rehearsal of 
talent, and (4} editing tempo. After producing nearly 100 training films, the author 
has a background to present the reason "why" for the distinct pattern followed in train- 
ing-film production. 

More than 1900 motion pictures for training purposes were made 
by industrial-film producers in four years ; an unprecedented era for 
the commercial motion picture. These 1900 subjects were written, 
produced, and released between 1941 and 1944. They were produced 
for training the military forces and training within industry, and do 
not include films made by Hollywood or the motion picture service 
divisions of the military forces. 

It was proved that the use of visual methods speeded up training 
as much as 34 per cent, as well as doing a more thorough job. During 
our war years, as never before, motion pictures helped do a job faster 
and better, and were a contribution, in no small way, to the war effort. 

Because there has been a continued use of training, educational, 
and documentary films throughout the world, the discussion of prob- 
lems in production appears timely. 


In a training film, the minimum number of sets should be used. 
It is unnecessary to motivate the action by location changes as in the 
entertainment picture. The training film should be a presentation 
of one subject, treated in logical sequence, with one simple location 
or set, if possible. To cover too much subject matter in one film is 
dangerous. The trainee can retain the technique much easier if it is 
presented in an unpretentious manner. 

The sets must be authentic. There is no need for highly stylized, 

* Presented Apr. 23, 1947, at the SMPE Convention in Chicago. 
** Reid H. Ray Film Industries, Inc., St. Paul 1, Minn. 


204 RAY Vol 49, No. 3 

dramatic, or tricky sets for photographic effects. If a training film 
on how to operate a horizontal milling machine is to be made, the set 
should look like a real machine shop. Nothing should be included 
in the set to distract the attention, of the trainee while viewing the 
film. The trainee may have spent several years in a machine shop 
as an apprentice and if the film is to "teach", it has to have every ap- 
pearance of genuine authority. The film must show as well as speak 
the language of the machine shop. A training film needs no enter- 
tainment to do its job thoroughly. It is a good practice to visit 
several factories, before constructing sets, to be sure that the set which 
will appear on the screen will look like a real shop, and not a motion 
picture set. Often a set can be built in a factory around the machine 
to be used as the "prop" for the film. In these instances simplifi- 
cation of background is important. 

In a series of 20 electrical training films, "live" sets had to be used. 
These sets represented new and old houses which were to be wired. 
In the "new-house" wiring films, the rooms of the house were in the 
unfinished building stage. There were two-by-four room partitions, 
rough flooring, siding, and cement building blocks for foundations. 
In the old-house sets, there were finished walls complete with lath, 
plaster, and wall paper. The two by fours were 16 inches on center, 
and the headers in the walls were properly spaced, because the prob- 
lems were those actually found on electrical-installation jobs. 

It was necessary to build a full-scale, two-story, one-room house in 
the studio for the films on wiring old houses. It was quite impossible 
to show authentically the work required in a one-story conventional 
set. One sequence illustrated the method of tearing up the upstairs 
floor and running cable under the floor for the first-floor ceiling fixture 
There, again, the set had to have lath, plaster, rough and finished floor 
ing complete; yes, and even old-style wall paper used 20 years ago. 

A similar technique should be followed in selection of props and 
"gadgets used in the film as "workpieces". In machine-shop work, 
genuine parts should be used, not a demonstration workpiece that 
never ends up being a recognizable tool or part. If necessary, 
simple part was designed, but an actual part that could be machined 
to illustrate the teaching problem usually was found. A training 
film should never be just an "exercise" or demonstration. The 
trainee should be shown an actual job in process. 

In several films produced for the Navy on instrument control it: 
the flying of aircraft, there were problems to solve. In 


entertainment field of motion picture production, considerable liber- 
ties are taken in presenting simulated flying conditions, aircraft con- 
trol while in flight, and actual cockpit conditions. Our training films 
were to be used by advanced groups, men who had gone through 
their primary training and were ready for advanced procedures. A 
film that did not measure up to real-life conditions would have been 
laughed off the screen. To the technical advisers the Navy assigned 
to commercial producers, we are gratefully indebted. These men gave 
valuable assistance. If an SNJ aircraft was to be flown, the close-ups 
in the cockpit were made in an SNJ even though it worked con- 
siderable hardship on the camera crew. When aerial shots had to 
be made with the aircraft in the studio, we were most particular 
with the resulting simulated effect. Smoke bombs, mineral oil, and 
spray guns created an authentic overcast. 

Strict attention was given to all control movements in the cockpit. 
For example, every movement of the "stick" to control the move- 
ment of the aircraft was carefully rehearsed so that the amount of 
movement corresponded to that used in actual flight. 

When flying in overcast weather, considerable moisture accumu- 
lates and runs down the front of the cockpit windshield. The mineral 
oil gave that same desired effect and created a misty windshield. 
Flying scenes were made in overcast weather, with both the camera 
plane and the subject plane flying in and out of the overcast. This 
was the type of weather trainees were expected to experience, so it 
was made as genuine as possible. 


In a recent Hollywood production, "Lady in the Lake", produced 
and directed by Robert Montgomery, training-film camera technique 
has been employed. The camera is used as the narrator, and the 
audience views the story through the eyes of the storyteller. One of 
the important things in making training films is to choose a camera 
angle which is called the "operator's viewpoint". The scene is shot 
from as near an angle as the operator views the controls, the work- 
piece, or the reading of a micrometer. In the usual factual film, the 
camera angle is selected for good photography, or perhaps an im- 
pressive low-angle shot. Not so with the training film. The trainee 
must see the picture as if through the eyes of the man operating the 
machine. The camera must have that viewpoint. 

206 RAY Vol 49, No. 3 

Very often this imposes a difficult assignment on the cameraman, 
not only in camera placement, but in lighting. But it pays off in 
good training-film technique for the trainee sees a control wheel 
actually turned left, not in the opposite direction as would be the case 
if it were photographed facing the operator of the machine. 

In many training films, especially those on precision-measurement 
work, we found that the usual extreme close-up was not close enough. 
A new photographic word was coined to mean an ultra-extreme close- 
up, and that word was "macrophoto". The standard lenses on the 
Mitchell and Bell-Howell cameras used would not rack out far enough 
to give this close a close-up, so Howard Cress of our Camera Depart- 
ment, turned out some lens-extension tubes. Thus, close-ups of 
great magnification were obtained with unusual photographic effects. 
For example, to show a small "burr" on the side of an aluminum pump 
block, a macrophoto was made. The small "burr" was approxi- 
mately one-half inch in width but in the close-up, this "burr" filled 
two thirds of the plate. 

Many times these- enormous close-ups were used when micrometer 
readings were to be taken. In such close-ups, we were able actually to 
show the readings on the barrel of the "mike". So with the training 
film, came a new word in motion picture language, the macrophoto. 

Selection of a camera angle is most important when measuring or 
checking dimensions with a dial indicator. The usual method is to 
shoot the workpiece being measured, and then photograph an ex- 
treme close-up of a dial set to the correct dimension and intercut 
these two shots. Whenever possible the camera angle was arranged 
so that both the workpiece to be checked and measured and the dial 
indicator were in the shot. The trainee was sure that the measure- 
ments were being taken on the actual workpiece. This often called 
for a change of focus and some rather odd setups, because the dial 
indicator might be 10 or 11 inches from the plane of focus of the 
measuring device. Other times a short pan or tilt-up to the dial 
would suffice to convince the trainee that no camera tricks were em- 
ployed. The method we used was far more acceptable to the tech- 
nical men in the business. 


The usual cast for a factual and documentary film is composed of 
professional actors. This is not possible in training films which deal 


with specialized fields and talents. Experts were employed in the 
specialized field the training film embraced, and schooled in motion 
picture acting. 

There is, at the beginning of the shooting schedule, a slight delay 
in having to instruct the mechanic, electrician, foundryman, or drafts- 
man in motion picture technique. But within a day or two at the 
most, shooting schedules are speeded up using the man who "knows 
his business". These technical men often worked with the writers 
when the process to be photographed was put into script form. 
Therefore, they knew the entire film story and there was a double 
check that the scene was all right after the director called "Cut". 

In the usual training film, 75 per cent of the footage is close-up and 
the face of the character is not seen. The hands and what those 
hands do are most important. It is the manner in which a skilled 
mechanic's hand picks up tools, turns dials, and handles workpieces 
with a sureness and confidence that brings to training-film photog- 
raphy an art that a professional motion picture actor can never learn. 
That is logical because the machinist may have spent a lifetime 
doing just what is being photographed. 

In using talent skilled in the field of the training-film subject, the 
possibility of retakes is also minimized. If careful judgment is used 
in the selection of camera angles and the technician approves the 
action of a take, there will be little argument when the "rushes" are 
screened. No expensive retakes will be necessary as sometimes 
happens after the rushes are screened. 

To retake some scenes would have meant delays and added 
expense. One script called for an electrician to tear up a floor to pull 
in armored cable. This called for skill and know-how, without benefit 
of rehearsal. Another bit of action required the electrician to remove 
a section of wall paper, then cut through plaster and lath. Here, a 
wrong movement would have required redoing and patching up the 
wall, waiting for the plaster to dry out, an expensive delay. The 
electrician-actor employed was an old hand at these tricks, he did his 
work swiftly, accurately. Many of these scenes would not allow re- 
hearsal of the complete action, only sketchily outlining the action for 
camera lines. These were really "one-shot" takes. 

To achieve such results, co-ordination between the technician, the 
director, and the cameraman is required. It can and has been done 
when attention is given in selecting talent for training-film production. 

208 RAY 


It must be remembered that the training film does not entertain, 
but shows why, how, when, and what for, and therefore editing tempo 
must be considered. Entertainment films are edited according to 
type; fast comedy, heavy drama, action pictures; all have a tech- 
nique of editing for tempo. 

In editing a training film, there should not be cuts that are confus- 
ing as to orientation of the trainee's viewpoint as he watches a ma- 
chine process. Generally, scenes should run long; short three- or 
four-foot scenes are confusing, unless the same angle has been shown 
before for orientation. 

Selecting one take, and running that take through a ''cycle" is 
better than trying to get a variety of angles. In training films one 
good camera angle of a cycle is better than trying to edit the film so 
that several angles can be intercut. 

The usual factual or documentary film uses narration on about 90 
per cent of the footage, the other 10 per cent is music, effects, or runs 
silent. The properly produced and edited training film should have 
about 45 per cent of the footage with narration; 55 per cent of the 
footage silent, as we know that 76 per cent of all impressions are made 
through the eye. Off-stage voice is used to guide the thinking of the 
audience, or to point up a certain action for emphasis. Some training 
films have scenes running 120 to 150 feet in length with only forty or 
fifty words of narration. "Let the picture tell the story" is a good 
maxim for good training-film editing and scoring. 

Slightly off the scheduled subtopics of this paper is this added 
thought : the type of scenario for shooting a training film. The most 
important part of that script is the "action continuity". There must 
be smooth, concise, and complete continuity of each sequence. The 
narration included with the "shooting script" can be tentative for it 
is usually revised several times before the picture is recorded. A 
script with complete "action continuity" is an aid to be sure that the 
picture tells the story. 

For the many hundreds of training films to be made, perhaps these 
comments will help produce films that will do a better job training 
new technicians for our mechanized world. 



Summary. The theory of light production in the high-intensity carbon arc is dis- 
cussed, together with a description of the phenomena associated with the initial striking 
of the arc and the maintenance of the electric discharge through the arc stream. The 
formation of the positive carbon crater is described and the factors defined which deter- 
mine the maximum current loading which a particular carbon electrode will support. 
The importance of efficient heat dissipation from the positive crater region in extending 
the useful current range of a given-sized carbon is pointed out, and the effectiveness of 
water cooling in providing better heat dissipation at this point is noted. 

In the motion picture industry, light is perhaps the most important 
single factor in the recording and reproducing processes involved. 
Light is thrown in carefully controlled quantities and distribution 
patterns on the motion picture sets and on the actors. Reflected por- 
tions of this incident energy are directed toward a camera lens, and 
focused to produce a permanent record of this reflected pattern on 
film. Finally, light is selectively absorbed by prints from this film in 
a theater, so that the distribution pattern reaching the original film 
in the studio camera may be recreated on the theater screen. Light 
is also an essential agent in the recording and reproduction of sound, 
although that important phase of the industry will not be included in 
this present consideration. 

The high-intensity carbon arc is such a commonplace and generally 
useful light source in the photographic and projection processes which 
characterize the motion picture industry 1 ' 2 that little thought ordi- 
nar.ily is given to the physical processes involved in the operation of 
such a source. In the belief that concepts found useful in the labora- 
tory in directing the development of new and brighter carbons may 
be of interest to the ultimate users of such carbons, the present dis- 
cussion has been prepared. 

To begin with, consider the simple arc circuit of Fig. 1. Here a 
direct-current source of perhaps 110 volts is connected through a 
series resistor to a pair of carbons. In common with all gaseous dis- 
charges, the carbon arc has a negative resistance coefficient: as the 

* Presented Apr. 22, 1947, at the SMPE Convention in Chicago. 
** Research Laboratories, National Carbon Company, Inc., Cleveland 1, Ohio. 




Vol 49, No. 3 

current is increased, the ohmic resistance of the arc becomes less. 
Some ballast effect, such as is provided by the series resistor in this 
case, must, therefore, be incorporated in the circuit. To start the arc, 
the two carbon electrodes are brought into brief contact, drawn apart 
again, and a light source of very high intensity is produced where 
nothing but air existed before. 

When the power is first applied, nothing happens, because the cir- 
cuit includes an air gap between the carbon electrodes which cannot 
be broken down by the relatively low voltage of the power source. 
It is not until the electrodes are brought into physical contact that 
current starts to flow. In a series circuit, such as is established when 
the electrodes touch, the same current flows throughout, so that the 


FIG. 1. A typical carbon-arc circuit. 

relative heat in any portion of the circuit is determined by the resist- 
ance of that portion. At the point of contact between the electrodes, 
the cross-sectional area is small, so that the resistance is high, ^.s a 
result, a high concentration of heat is produced at this point; and as 
the pressure on the electrodes is reduced, preparatory to separating 
them, the contact area grows smaller and smaller, so that it continues 
to become hotter and hotter. 

In order to explain what happens next, it is necessary to consider an 
atomic property of hot bodies called thermionic emission. The 
atoms in any solid substance are in a continual state of vibration, with 
electrons revolving rapidly around each nucleus in a variety of orbits 
at various distances. When the substance is heated, the atomic 
vibrations grow more intense, and the electrons spin faster and faster 
through wider and wider orbits. In the case of those atoms next the 


surface, an occasional electron will break away altogether, and as the 
heating continues more and more electrons fly off into space. 

Hot carbon is not so good an emitter of electrons as the materials 
used for this purpose in vacuum tubes, but it does possess this prop- 
erty to an appreciable degree. Thus, to return to the arc, as the last 
pair of atoms is about to be drawn apart in the separation of the elec- 
trodes, the concentration of current and the intensity of the resultant 
heat are terrific, sufficient not only to cause thermionic emission, but to 
vaporize the carbon itself at the tiny area of final contact. Consider 
also what would happen at the instant of separation if no arc were to 
form, so that the current would fall abruptly to zero. The full open- 
circuit line voltage would immediately appear across the gap, and if 
the initial gap is assumed to be a millionth of an inch, and the line 
voltage 100 volts, a voltage gradient of 100,000,000 volts per inch 
would be established promptly. As a matter of fact, the distance be- 
tween atoms in solid carbon is something of the order of 20 billionths 
of an inch, and it might be assumed that at the instant contact is 
broken, the physical separation is of this magnitude, which would give 
a voltage gradient of five billion volts per inch. 

It naturally follows that such a tremendous voltage gradient, acting 
in combination with the free electrons around the white-hot electrode 
tips, causes enough of them to flow across the gap as the carbons are 
pulled apart so that current continues to flow through the hot gases, 
and an arc is established. In turn, the high concentration of power 
within the narrow confines of the arc produces light, after the follow- 
ing fashion : 

To begin with, there is incandescent carbon at its volatilization 
temperature of over 6500 F (3600 C). Since a temperature of only 
2600-2900 F (1425-1595 C) is enough to produce a "white heat", it is 
apparent that incandescent carbon alone is responsible for a good 
share of the brightness of the carbon-arc crater; for all of it, as a 
matter of fact, in the low-intensity carbon arc, and from a fifth to a 
half of the total brightness in common high-intensity trims. 3 

The increased brightness of the high-intensity arc is the result of the 
combination of a high current density, i. e., a high concentration of 
electrons in the arc stream, and an atmosphere in the positive crater 
region rich in "flame materials" volatilized from the special coring of 
the positive electrode. These flame materials are in most cases com- 
pounds of the cerium group of rare-earth metals, combined in a mix- 
ture with carbon in the core. As the carbon shell burns away to form 



Vol 49, No. 3 

a crater, as indicated by Fig. 2, the core is exposed to the extreme arc 
temperature and is vaporized into the crater enclosure. Here, the 
rare-earth particles are bombarded by electrons to produce very in- 
tense light. It is perhaps helpful here to picture a maelstrom in the 
positive crater, with many billions of rare-earth atoms continually 
colliding with as many electrons. As the result of each collision, a 
rare-earth atom absorbs energy from an electron, and is transformed 
into an "excited" state. In other words, the excited atom possesses 

FIG. 2. Diagram showing the mixture of rare-earth 
atoms and electrons in the positive crater of the high- 
intensity carbon arc. 

an amount of energy in excess of the normal stable value. Moreover, 
as'defined by quantum theory, this excess energy may have any one of 
a number jof discrete values, depending upon the number and arrange- 
ment of the electrons circulating around the atomic nucleus. The 
rare-earth atoms have many electrons (cerium has 58, circulating in 14 
different orbits) so that the likelihood of scoring a hit on such a large, 
well-populated target is correspondingly increased. At the same 
time, there are a great many excited states possible, so that the likeli- 
hood is excellent that a hit will produce excitation. Since the rare- 
earth atoms are not stable in tnese excited states, they immediately 

Sept. 1947 



give up their excess energy. This they do in the form of pulses of 
radiation, each having a particular wavelength associated with the 
excited state ; or an excited atom may return to a normal energy level 
in a series of discrete steps, emitting radiant-energy pulses of as many 
different wavelengths on the way. It is characteristic of the rare- 
earth atoms that these energy pulses are of wavelengths to which the 
human eye is sensitive, and that they are distributed in such great 
numbers over the range of visual sensitivity that an essentially equal 
energy spectrum or a "white" light is produced. In this way, the 
brightness of the high-intensity carbon-arc crater is increased many- 
fold over that of the plain car- 
bon arc (to over ten times, in the 
laboratory) . 

Fig. 3, which shows a picture of 
a typical high-intensity carbon arc, 
can now be viewed with a new un- 
derstanding of what is going on. 
From the incandescent tip of the 
negative carbon underneath, count- 
less numbers of electrons are being 
drawn out into the arc stream and 
accelerated like bullets toward the 
positive electrode by the voltage 
gradient along the arc stream. To 
make enough electrons available, 
i. e., 63, followed by 17 zeros, elec- 
trons per second for each ampere, 
the negative tip must be heated to 
a very high temperature, hence the 
bright tip and red heatback of this 
electrode. These electrons rush 
across the arc stream, meeting 
nothing much except air atoms 
until they approach the region of the positive carbon, a bluish light 
resulting from collisions with the air atoms in the arc stream. At the 
crater, and particularly inside it, the electron stream encounters the 
rare-earth atoms, with a resultant production of brilliant white 
light. Under the influence of convection currents established by 
the hot gases, a bright stream of excited rare-earth atoms emerges 
from the crater and drifts upward into the tail flame. 

FIG. 3. The inclined trim high- 
intensity carbon arc. (13.6-mm posi- 
tive carbon at 150 amperes with the 
negative directed upward at an angle 
of 53 degrees from the horizontal.) 

214 BOWDITCH Vol 49, No. 3 

Carbon is an ideal material from which to construct the electrodes 
for such an arc because of three important properties : (a) it is a good 
electrical conductor, (b) it remains in solid form to a very high tem- 
perature (approximately 6500 F (3600 C)), and (c) it volatilizes di- 
rectly without passing through a messy molten state. 4 

The positive electrode of the high-intensity carbon arc differs from 
that employed with the low-intensity arc in two important respects. 
The core is not only much larger, but it is also heavily loaded with the 
flame materials whose light-producing function has just been de- 
scribed. In low-intensity positive carbons, the core hole is no more 

FIG. 4. Comparison of the core sizes and craters of typical low- and high- 
intensity carbon arcs. (The deep high-intensity crater shown in the upper 
cross section was formed on a 13.6-mm carbon at 150 amperes, the almost 
flat low-intensity crater on a 12-mm carbon at 30 amperes.) 

than one fourth the diameter of the shell : in the high-intensity posi- 
tive carbon, the core is at least one half and is frequently a much 
greater proportion of the outside diameter of the shell. This is illus- 
trated by Fig. 4. The current density is also much higher, a one-half 
inch low-intensity carbon operating at about 35 amperes as compared 
with well over 100 amperes for the same-sized high-intensity positive 

Because of the lower voltage drop from the arc stream to the core 
as compared with the voltage drop to the carbon shell, most of the 
electrons forming the high current in the arc stream are encouraged to 
travel to the central core. Here the concentration of energy is so 
great that the core and the immediately surrounding shell are vapor- 
ized faster than the shell at the outside. Thus, as shown by Fig. 4, a 

Sept. 1947 



cup or crater is formed on the end of the positive carbon, which is 
filled with the rich light-producing mixture of rare-earth vapors and 
electrons. As the current is increased, the depth of this crater like- 
wise increases to a limiting value determined by what is called the 
"overload" of the carbon. Overload is characterized by the fact that 
beyond a particular current value the arc no longer burns smoothly 
and quietly, but becomes unsteady and noisy. Since, for all important 
uses, the arc must be both stable and quiet, operation is always con- 
fined to currents well below this overload value. 


FIG. 5. Diagram illustrating the mechanism of 

overload in a high-intensity carbon arc. 
AB = normal electron path to core 
AC = overload electron path to carbon shell. 

As shown in Fig. 5, the mechanism of overload is visualized as fol- 
lows. Since the electrons encounter a much lower voltage drop in en- 
tering the positive carbon through the central core, the resulting 
crater becomes deeper and deeper with increasing current. At the 
same time, the tendency for electrons to flow directly to the shell C at 
the mouth of the crater, instead of taking the longer path to the bot- 
tom B, becomes greater and greater. Finally, the difference between 
the voltage drop over the longer arc stream AB to the bottom of the 
crater, and the voltage drop along the shorter path AC to the crater 

216 BOWDITCH Vol 49, No. 3 

lip, becomes sufficient to counteract the effect of the more favorable 
electron entrance into the core. When this happens, the electrons 
travel in increasing numbers to the shell. Here the rate of energy re- 
lease increases to a point where carbon is volatilized violently and 
noisily. Thus, an upper limit of about 35 volts (the anode drop to 
pure carbon) seems to be imposed on the voltage component of the 
energy which can be released within the positive crater. The current 
component of this energy is less rigidly limited. For instance, water- 
cooled positive jaws may be employed, with a minimum protrusion 
of the positive carbon beyond these jaws. 5 The more efficient heat 
dissipation thus obtained permits the use of substantially higher cur- 
rents and the achievement of correspondingly higher brilliancies. 
The added complication of arc operation with water-cooled jaws has 
so far prohibited their use in many applications. However, in motion 
picture studios where background projection is frequently employed 
to provide the setting for a physically distant location, operation with 
water-cooled jaws to achieve the brightest possible background image 
is receiving active experimental consideration. 

It is hoped that these theoretical considerations, assembled in the 
course of arc-carbon research and development, have proved of inter- 
est outside that limited field. Not only the designers and the opera- 
tors of the many types of burning mechanisms which facilitate the 
generation and release of this benign form of atomic energy, but also 
the many artisans engaged in the control of this energy to create 
wanted effects on film and on the motion picture screen, are all de- 
pendent upon the radiant output of the carbon arc. It is to them that 
this paper has been directed. , 


1 KALB, W. C., "Progress in Projection Lighting," /. Soc. Mot. Pict. Eng., 35, 
1 (July 1940), p. 17. 

2 LINDERMAN, R. G., HANDLEY, C. W., AND RoDGERS, A., "Illumination in 
Motion Picture Production," J. Soc. Mot. Pict., 40, 6 (June 1943), p. 333. 

3 MACPHERSON, H. G., "A Suggested Clarification of Carbon Arc Terminology 
as Applied to Motion Picture Industry," /. Soc. Mot. Pict. Eng., 37, 5 (Nov. 
1941), p. 480. 

4 CHANEY, N. K., HAMISTER, V. C., AND GLASS, S. W., "Properties of Carbon 
at the Arc Temperature," Trans. Amer. Electrochem. Soc., 67, (1935), p. 201. 

6 JONES, M. T., ZAVESKY, R. J., AND LOZIER, W. W., "A New Carbon for In- 
creased Light in Studio and Theater Projection," /. Soc. Mot. Pict. Eng., 45, 6 
(Dec. 1945). p. 449. 



MR. DOFFER: The speaker mentioned that carbon is a very good electrical 
conductor, therefore you had a low loss. Why do some carbons have a copper 
coating on them? 

MR. F. T. BOWDITCH: Carbon is not so good an electrical conductor as copper. 
I didn't mean to imply that. But as compared with other materials which remain 
solid at a comparable temperature, carbon is an excellent conductor. A copper 
coating is applied to carbons in services where the current must be carried through- 
out the length of the carbon from fixed clamps at the end farthest from the arc. 
Bare carbons are ordinarily employed with current jaws which grip the carbons 
close to the arc, and through which the carbon is pushed forward by a feeding 
mechanism at the rear. 

MR. READ: Does the alternation of the electron path from points A-B to 
points A-C take place about once a second or does it occur at a rate which pro- 
duces an audible note? 

MR. BOWDITCH: I did not mean to imply there was any alternation whatever. 

MR. READ: I don't mean alternating current. 

MR. BOWDITCH: The picture we have, and the picture I intended to portray, 
is that at reasonable currents, well below the overload limit, practically all the 
electrons take the longer path to the base of the crater. This is so because it only 
requires an anode drop of perhaps five volts to transfer electrons from the arc 
stream into the core at this point, as compared with 35 volts directly to carbon at 
the crater lip. Now as the crater becomes deeper, more and more electrons con- 
sistently take the shorter path to the crater lip (not first this way and then that 
way), until a sufficient electron concentration is developed to generate a very in- 
tense heat at that point. Every electron entering the carbon releases 35 electron- 
volts of energy as it lands there, and so is responsible for a very high concentration 
of energy. 

MR. O. E. MILLER: Do you have positive-ion current in appreciable amount? 

MR. BOWDITCH: Our measurements indicate that the positive ions carry only 
a small percentage of the total current. There is such an effect. However the 
energy relationships indicated by the crater radiation spectrum show that only a 
very small percentage of the rare-earth atoms are ionized. Positive-ion current, 
while present, is therefore considered to be a minor factor in the light-producing 
relationships involved. 


M. T. JONES** 

Summary. This paper describes a theoretical method for determining motion 
picture screen light in a carbon-arc projection system, employing light measurements 
made directly on the arc crater from various angles of view. The opportunity is thus 
afforded to study the performance of a given carbon-arc source in optical systems having 
a much wider variety of collecting angle, magnification ratio, and optical speed than 
could possibly be assembled for actual optical-bench tests. In this way, it is hoped 
that the selection of the best possible optical system for a given arc will be facilitated, 
and vice versa. 

Examples of the application of the method to the 8-mm "Suprex" carbon arc at 70 
amperes, and to the new 13.6-mm super-high-intensity carbon arc at 290 amperes are 
given. In each case the analysis has been extended for illustrative purposes over a 
wider range of optical systems than is of present practical interest. However, the 
general conclusions thus made possible were considered of sufficient theoretical interest 
to justify this otherwise impractical consideration. For instance, it is indicated that, 
particularly with the larger crater, the greater lumen pickup of a high collecting angle 
does not always result in more light on the screen. 

Such practical checks of the method as have been made to date, comparing screen 
light measurements in complete optical systems with the values predicted by the 
method, indicate a useful order of accuracy for optical-design purposes. However, 
the limited scope of these checks is recognized, as is the fact that the calculations do not 
take screen color variations with focal position into account. 

It is hoped that this paper will stimulate others to carry on this same type of analysis, 
and that the foundation will thus be laid for the most effective combination of carbon 
arcs with the associated optical systems. 

In practically every 35-mm motion picture projection system, the 
light on the screen originates in the crater of a carbon arc. It is col- 
lected by a mirror or condenser and focused on a picture held in the 
film gate; the picture thus illuminated is imaged on the screen by the 
projection lens. An important part of the study of a carbon arc for 
this service is the evaluation of the light output from the arc crater in 
terms of the screen illumination produced after the light passes 
through such a chain of optical elements. The method of evaluation 
described herein is based upon measurements of the brightness distri- 
bution over the arc -crater region from various angles of view, and 
the mathematical integration of these measurements to yield aperture 

* Presented Apr. 25, 1947, at the SMPE Convention in Chicago. 
** National Carbon Company, Inc., Fostoria, Ohio. 


illumination data. In this way, the performance of a particular car- 
bon arc in a wide variety of systems of different optical speed, magni- 
fication ratio, and pickup angle can be predicted. This reduces the 
time required for such an analysis to a small fraction of that required 
to cover this same range by actual assembly and over-all measurement 
of the many individual systems. Such a procedure thus contributes 
importantly to the determination of the best possible combination 
of carbon arc and optics to satisfy a particular projection requirement. 
In the present analysis, ellipsoidal mirrors are the only light-collecting 
elements directly treated, although it is shown that the procedure so 
developed can be applied with reasonable accuracy to condenser op- 
tics as well. 

The characteristics of the optical system are such that each point 
on the screen receives light from many points in the arc. The result- 
ant illumination is thus a complex function of the brightness distribu- 
tion over the arc region. It is possible to measure this brightness dis- 
tribution with considerable accuracy. 1 The next step in a calculation 
of screen light is therefore an analysis of such brightness data in com- 
bination with the optical properties of an ellipsoidal mirror. Both the 
manner in which light is collected from various angles of view of the 
crater region and the patterns in which this light is focused on the 
film gate must be considered to give the amount and distribution of 
the illumination there. Finally the action of the aperture, projection 
lens, shutters, heat filters, and draft glass in reducing the luminous 
flux and in changing its distribution must be considered to determine 
the light on the screen. 

The light distribution projected on the film gate by an ellipsoidal 
mirror is calculated by a method similar to that developed by Benford 
for the distribution of light in a searchlight beam projected by a parab- 
oloidal mirror. 2 ' 3 However, in order that the process might find 
more extensive experimental use, it was necessary to develop many 
simplifications which permit rapid evaluation of a given arc without 
significant loss in accuracy. The procedure so developed is based 
upon the annular symmetry of the optical system. That is, all 
points in the mirror which lie on a circle centered on the axis, form 
crater images on the film gate which, for all practical purposes, are of 
the same size, shape, and brightness distribution. It is thus advan- 
tageous to treat each element of the system (the crater region, the 
mirror, and the film gate) in a series of concentric annular regions. 
The light contribution of each such region can be determined with 

220 JONES Vol 49, No. 3 

comparative ease, for summation with others similarly obtained. 
The method, then, is developed along the following lines: 

(1) The film gate area is divided into a series of concentric zones of 
equal width, centered with the film aperture. 

(2) The mirror surface is divided into a series of concentric seg- 
ments of such width that all contribute alike to the illumination per 
unit area on the film gate. These segments must decrease in area with 
increasing angle of view, to compensate for the accompanying de- 
crease in image-magnification ratio. 

(3) Crater-brightness-distribution data are recorded from the an- 
gles of view corresponding to the effective centers of these mirror 

(4) Each projected crater area so explored is divided into a series of 
concentric zones of such width that, in combination with the average 
magnification ratio of the associated mirror segment, the crater-zone 
images at the film gate will exactly match the zones established there. 

(5) The contribution of each mirror segment to the illumination of 
a particular zone on the film gate is determined from the crater-bright- 
ness values recorded over the associated zone at the crater. A summa- 
tion with similar contributions of the other mirror segments gives the 
total illumination of that film-gate zone, and the similar treatment of 
each film-gate zone in turn gives the total illumination of the film gate. 

(6) Finally, a determination of the film-gate illumination which en- 
ters the film aperture and the effect of the projection lens on it give a 
measure of incident screen light. 

In brief, this is the method. A more detailed description requires a 
consideration of the fundamental properties of the ellipsoidal mirror. 


The surface of an ellipsoidal mirror is generated by the rotation of 
an ellipse about its major axis, A- A' of Fig. 1. Such a reflector pos- 
sesses two symmetrical focuses, FI and F 2 , so related that light 
emitted in any direction from either one is reflected to the other. Ac- 
cordingly, as used in service, the arc crater is located facing the mirror 
at one focus F\ and the film gate is placed at the other. The entire 
ellipsoid, shown at the top of Fig. 1, is never employed in such a mo- 
tion picture projection system; only the section from P to P', shown 
separately below, is needed to fill the projection lens. 

Since the carbon-arc crater must have a small but finite area in 


order that its image on the film gate will be large enough to cover the 
aperture, light will originate.not only from the true focus FI, but from 
near-by off -axis points as well. When the paths of these off -axis rays 
are calculated from the geometry of the ellipse, it is found that each 
elemental area of the mirror focuses an image of the crater centrally 
upon the aperture. These images vary in shape and in size, depending 
upon the angle from which each elemental mirror area looks at the 

A 1 



FIG. 1. Focusing action of ellipsoidal mirror. 

crater. Referring to the lower drawing of Fig. 1, the size and shape 
of each image is related with this viewing angle 6 (at the left) in the 
following way. As 6 increases, the size of the image decreases. This is 
so since the magnification for any point 5 on the mirror is equal to the 
distance F 2 S, divided by the distance FiS; and it is apparent that 
the distance from mirror to film gate F 2 S decreases as 6 becomes larger. 
Also, the circular crater of the carbon arc appears to be an ellipse as 
viewed from any off-axis point S, so that an elliptical image is formed 
at F%. The dimensions of this ellipse are 

major axis (FzS/FiS) X (crater diameter) 
(F 2 S cos 0) 

minor axis 

cos 8') 

X (crater diameter). 

222 JONES Vol 49, No. 3 

For simplification of later calculations, cos 8' is assumed to be unity, 
since the angle 6' is always sufficiently small that no serious error is in- 
troduced iathe results. (At//2.0, cos0' = 0.97; at//1.4, cos0' = 0.94.) 

The variation in image with viewing angle 6 is illustrated in Fig. 2, 
which shows arc images as they would be formed on a film gate for 
four different values of 6. The arc is the 13.6-mm National High 
Intensity Projector Positive operated at 150 amperes. For compari- 
son, the outline of the standard 35-mm sound projection aperture has 
been superimposed, with its size adjusted to the axial magnification of 
3, characteristic of the //2.2 condenser system frequently employed 
with this arc. While this system has a maximum value of 6 = ap- 
proximately 40 degrees, the images out to 9 = 75 degrees are shown to 
illustrate the degree of coverage obtained at the wider angles com- 
monly employed with mirror systems. This is done for illustrative 
purposes only, since it is not practicable to employ an arc of this 
power with present-day mirror systems and since such a system would 
be much faster (//l.O) than can be utilized in practice. While the for- 
ward crater image is much larger than the aperture (thus providing 
what appears to be an excessive waste of light), the crater image for an 
angle of 6 = 75 degrees covers only a small portion of the aperture, 
because of the reduced magnification and the narrowness of the fore- 
shortened view at this angle. 

The image of the arc which would be formed by a complete mirror 
is a composite of many individual images, samples of which are shown 
in Fig. 2. The calculation of light passing the aperture must consider 
all these images. Moreover, light passing the aperture from images 
with large values of 6 includes some light from in front of the crater, 
some from the crater itself, and some from the carbon shell behind the 

As long as the center of the crater is maintained at one focus of the 
mirror, the centers of all images will coincide with the center of the aper- 
ture, as shown in Fig. 1 . This is the condition at which maximum light 
is projected on the screen. Accordingly, the present treatment has 
been confined to a consideration of the "in-focus" condition although 
it is possible to apply a similar treatment to "off-focus' ' conditions. 

There are three fundamental constants which define the optical ac- 
tion of a given mirror. Using the designations of Fig. 1, these are 

(1) the angle of collection 26 p 

(2) the speed (/number) 1/2 tan P ' 

(3) the nominal or axial magnification (Mo) 


FIG. 2. 13.6-mm crater images projected upon aperture by zones of ellip- 
soidal mirror at angles of view to the side- of the arc. 0.600 X 0.825-in. 
aperture. M = 3. 



Vol 49, No. 3 

The fixing of any two of these quantities defines the third, the rela- 
tions between them being plotted in Fig. 3. 

With this background, then, it is possible to proceed with the selec- 
tion of the film gate, mirror, and crater zones previously mentioned. 
The first and last of these will now be considered, leaving the mirror 
zones for separate consideration in an Appendix to this paper. 



a: 2.5 


i 1.5 




62.7 86.6 103.9 1175 1 2B.T 136.7 1472 1547 



10 I 

FIG. 3. Relations between speed, axial magnification, and collecting 
angle of ellipsoidal mirror. 


The manner of selection of annular zones of equal width on the film 
gate and on the crater is illustrated in Fig. 4. For illustration, this 
figure has been drawn for an axial magnification of 3, correspond- 
ing to the arc photographs in Fig. 2. The rings on the film gate have 
been drawn with a width of 3.0 mm requiring a width of 1.0 mm for 
the matching rings on the crater when viewed from 6 = degrees. As 
the angle of view increases to 25 degrees and to 75 degrees, the rings 
become wider, in direct compensation for the reduced magnification 
ratios at these angles. The portions of the crater region effective on a 
particular zone at the film gate thus vary widely with the angle of 


view from the mirror. For instance, each drawing of Fig. 4 has an 
equal number of zones, although the crater areas so covered are widely 


It is apparent that the crater-zone widths just described were fixed 
by the assumption of a 3-to-l magnification ratio, and that the con- 
sideration of systems of different magnification would require the 
calculation of an entirely different set of zone widths. Accordingly, 
in order to facilitate the treatment of a variety of optical systems of 


9 = 



FIG. 4. Annular zones on crater corresponding to similar zones 
on film gate. 

different magnification, it is advantageous to conduct the calculations 
on a proportional basis, introducing the magnifications of interest as a 
final step. To accomplish this, a hypothetical mirror of unit axial 
magnification is assumed, such that the proportional relation between 
magnification and viewing angle is the same as that of the average 
ellipsoidal mirror employed for motion picture projection. Fig. 5 
shows that the characteristics of a mirror of f/2.0 speed accurately 
represent the average mirror in this respect. Such a mirror is approxi- 
mately the average of mirrors with speeds from //1. 3 to //3.5, the 



Vol 49, No. 3 

range of greatest .practical interest. Even with a speed of //l.O, the 
error introduced by the assumption of the //2.0 characteristics 
amounts to no more than 5 per cent. 

For use with this proportional magnification system, a zone width 
of 0.5 mm on the film gate has been found to give results of satisfac- 
tory accuracy. The crater-zone widths from the axial viewpoint of 
= degrees then have this same dimension, and those at larger 
angles become progressively wider, in direct compensation for the 
reduction in magnification with angle. 


40 80 120 



FiG. 5. Relation between axial and edge magnification 
of ellipsoidal mirrors with various collecting angles and 

From the intensity distribution across the film gate calculated on a 
proportional basis in this way, the distribution for a particular mirror 
of interest is obtained in the following manner. Again, an axial mag- 
nification of 3 will be chosen as an example. In such a case, the film- 
gate zones will be of three times the width first calculated, or 1.5 mm. 
Therefore, the light will be distributed over nine times the area so that 
the intensities first calculated must be divided by that factor. 


The procedure so far described gives only the zonal illumination on 
the film gate. However, only the light which passes through 
the film aperture is of interest. Therefore, as shown in Fig. 6, it is 
necessary to determine that portion of the zonal illumination which is 
included within the rectangle. While this might be done by a rather 
tedious graphical procedure, apparent from an examination of Fig. 6, 
a much faster method is preferred. Such a method is described in the 
Appendix to this paper. 



In addition to describing the procedure for calculating mirror-zone 
widths and a short-cut method for determining that portion of the 
film-gate illumination which passes through the aperture, the Appen- 
dix carries through a typical calculation from actual arc measurements. 
The utility of the information so derived will now be discussed. 

Fig. 7 shows such data for the new 13.6-mm super-high-intensity 
positive carbon at 290 amperes. 4 The following conclusions are 
typical of those possible from such data : 

FIG. 6. Superposition of annular zones upon 
0.600- X 0.825-in. aperture when M = 5.4. 

(1) At a given speed, the side- to-center distribution ratio increases 
with collecting angle. 

(2) At a given collecting angle, the side-to-center distribution ratio 
decreases with increasing speed. 

(3) The luminous flux passed by the aperture does not increase in 
proportion to the light collected from the crater. With an//2.5 sys- 
tem, for instance, this flux varies no more than i 7 per cent from an 
average value of 45,000 lumens as the collecting angle is varied from 
60 to 140 degrees. With faster speeds, a more pronounced change in 
aperture flux occurs with change in collecting angle, with a peak value 
in all cases at some angle less than 140 degrees. 



Vol 49, No. 3 

(4) The side-to-center distribution ratio at a given optical speed 
increases consistently with increasing collecting angle. However, 
with a constant lumen system, such as the//2.5, the distribution ratio 
can only be increased by shifting light from the center toward the 
edges of the aperture. The higher ratios are thus achieved at the ex- 
pense of a lower central-light intensity. 

(5) If a given distribution ratio is desired, say 70 per cent, the 
luminous flux will be : 

45,000 lumens at//2.5 and 62 collecting angle 

70,000 lumens at//2.0 and 86 collecting angle 

100,000 lumens at //1. 6 and 112 collecting angle 

138,000 lumens at //1. 3 and 138 collecting angle. 






















. *"'" 



-- 1.6 







^ ^ 











1 6 






~ 2.5 




60 80 100 120 140 60 80 100 120 140 


FIG. 7. Analysis of light passed by aperture with various ellipsoidal 
mirrors and the new 13.6-mm super-high-intensity positive carbon at 
290 amperes. 

At //I.O, a distribution ratio no greater than 60 per cent can be ob- 
tained within the range of collecting angles investigated. Thus, as the 
speed is increased, the collecting angle required to produce a given dis- 
tribution ratio and the luminous flux increases. A larger crater size 
would be necessary to obtain a 70 per cent distribution ratio at //I.O. 
Also, where speed is limited, to//2.0 say, the data do not justify the 
use of a collecting angle higher than 90 degrees with this particular 
arc. It is also indicated that carbons smaller than 13.6 mm can be 
used with adequate aperture coverage. 


The 13.6-mm carbon just analyzed is not intended for use in mirror 
systems, but demonstrates the type of information available from the 
calculations described herein. The calculated performance of the 8- 
mm-7-mm "National" Suprex arc at 70 amperes shown in Fig.- 8 pro- 
vides an example of an arc intended for use with mirror systems. The 
curves are similar to those for the 13.6-mm carbon just discussed and 
similar conclusions may be drawn with the following comments. 

(1) The side-to-center distribution-ratio curves at a given speed 
now flatten with increasing collecting angle. 

(2) Because of the smaller crater size the side-to-center distribution 
ratio at a given speed is much lower than for the 13.6-mm carbon. 
The//1.0 and//1.3 speeds are not shown in Fig. 8 because the aper- 
ture coverage is far from adequate with these. 





























80 100 120 140 160 80 100 120 140 160 


FIG. 8. Analysis of light passed by aperture with various ellipsoidal 
mirrors and the 8-mm-7-mm "National" Suprex trim at 70 amperes. 

(3) Maximum luminous flux is reached at a collecting angle near 
130 degrees with no advantage indicated for employing collecting 
angles higher than this with f/2. 5 and//2.0 speeds. 

(4) The use of a lower collecting angle at these speeds would result 
in loss of light as well as in a lower distribution ratio with this carbon 
in contrast to the desirability for a lower collecting angle indicated for 
the 13.6-mm carbon previously described. 

The usefulness of the method for calculating the luminous flux 
which a given carbon will pass through an aperture under various 

230 JONES Vol 49, No. 3 

optical conditions is thus effectively demonstrated. As a tool for 
evaluating the performance of a carbon arc, this procedure should 
prove extremely valuable. 


In all the preceding considerations, no transmission losses in the op- 
tical system except those which occur at the aperture have been in- 
cluded. In order to translate the above-calculated luminous flux into 
light on the screen, it is necessary to take into account transmission 
losses such as mirror reflectivity, shadowing by the arc mechanism, 
heat filter, and draft glass, if employed, shutter, and projection lens. 
A treatment of this phase of the problem is beyond the scope of this 
paper. However, as a check on the validity of the calculations, sev- 
eral complete projection systems, both with mirrors and with con- 
densers as light-collecting elements, were assembled. Both screen 
light and brightness-distribution measurements were made. After 
making reasonable assumptions, based partially on actual measure- 
ment, for the transmission losses in the system, the screen light and 
distribution calculated from the crater-brightness measurements were 
found to agree quite well with the actual measurements. 

One of these checks was made with a standard 35-mm motion pic- 
ture projection system employing the National 13.6-mm High-Inten- 
sity Projector Positive at 150 amperes. With condensers operated at 
//2.0 and an //2.0 coated projection lens of 5-inch focal length, the 
maximum light on the screen with no shutter, film, or filters of any kind 
and with the standard 35-mm sound projector aperture, was measured 
to be 19,500 lumens at 60 per cent side- to-center distribution ratio. 
The results of calculating screen light from crater-brightness measure- 
ments are given in Table 1. 


Screen Light Calculated from Brightness Measurements on Crater of 13. 6- Mm 
High-Intensity Projector Positive at 150 Amperes 


Lumens Ratio, Per Cent 

Luminous flux passed by aperture with no 

correction for transmission losses 38,000 73 

Same including 69 per cent condenser trans- 
mission 26,200 73 

Screen light after applying losses in projec- 
tion lens 19,900 62 


These condensers operate with approximately an 80-degree collecting 
angle ; their measured transmission was 69 per cent. The transmission 
of the projection lens was measured to be 76 per cent with an 85 per 
cent side- to-center screen-distribution ratio from a uniformly illu- 
minated aperture. The agreement between calculated and meas- 
ured screen light values is within 2 per cent in this case. 

The fact that a check with a value based on mirror theory was 
obtained when condensers were used indicates the similarity at the 
smaller angles in the focusing action between these two types of 
light-collecting elements. Accordingly, application of the calcula- 
tions to simple condenser systems should be of use in evaluating car- 
bon arcs for this application. 


The writer wishes to express appreciation to Mr. R. J. Zavesky of 
the Fostoria Works, and to Mr. F. T. Bowditch of the Research Lab- 
oratories for many helpful and illuminating discussions in the course of 
the work. 



The mirror is divided into a series of concentric annular zones of 
such width that each is equally effective in contributing to the illumi- 
nation on the film gate. This choice simplifies the summation of crater- 
brightness data at all angles of view, in order to produce the illumina- 
tion on the film gate for the entire mirror. 

The determination of such 
a division of the mirror re- 
quires an integration over the 
mirror surface after the follow- 
ing manner . Referring to Fig . 
9, the intensity of illumination 
on a small area 5 on the mirror 
from a unit area spherical 
source of brightness B at FI, is 
B/R* where R is the distance from FI to 5. The total luminous flux 
per unit area of the light source over the annulus including S is equal 
to B/R 2 times the area of the annulus. Employing polar co-ordinates, 
the total flux </> M over the annulus whose edges are defined by m and 
6 n may be expressed as 

+ = ( Bn * p,*! sin Aft) U) 

Mm J<* 

232 JONES Vol 49, No. 3 

where the integral of the term in parenthesis is the area of the annulus. 
This simplifies to 

<t>M = 2rrB I n sin 6d0. (2) 

The total flux per unit area on the film gate <f> A reflected from this mir- 
ror zone is obtained by dividing </> M by the square of the magnifica- 
tion, which is the ratio of the image area on the film gate to the area of 
the source. The total flux per unit area on the aperture is thus 

0A = 

where M is the magnification at viewing angle 0. As previously dis- 
cussed, the magnification M varies with 6. The relation between M 
and B for any ellipse has been derived to be 

M = ki + & 2 cos (4) 


kl = M + * (5) 


In these equations M is the nominal or axial magnification of the 
mirror, and e is the eccentricity of the ellipse employed to generate 
the mirror surface. Substitution of Eq (4) in Eq (3) and integration 
produces the following expression for the luminous flux per unit area 
on the aperture from the mirror zone bounded by viewing angles 6 m 

and B n 

where M m is* the magnification at m , and M n that at B n . By sub- 
stitution of Eq (4) this gives 

(cos 6 m - cos n )] X B. (8) 

The expression in brackets in Eq (8) may be termed the "lumen 
factor" of a mirror zone, since it is the quantity by which crater 
brightness B must be multiplied to produce the luminous flux per unit 
area on the film gate. The "lumen factor" for the assumed "unit" 
magnification mirror is determined by multiplying Eq (8) by M 2 . 


The division of the mirror into annular zones of equal "lumen fac- 
tor" produces the desired zones of equal effectiveness at the film gate. 
With such a choice of mirror zones, crater-brightness data effective 
for any one zone can be directly averaged with data for all other 
zones. The lumens per unit area on the aperture for the entire 
mirror may thus be found by averaging the effective brightnesses of 
all the zones and multiplying by the "lumen factor" for the entire 
mirror. The equal -lumen zones listed in Table 2 have been calcu- 
lated for an//2.0 mirror, although, as previously indicated, these are 


Zones of Equal Lumen Factor for an f/2.0 Ellipsoidal Mirror with M 5.4 



Limits of Zone 






22 27' 

31 20' 


37 56' 

31 20' 

43 19' 


47 55' 

43 19' 

51 56' 


55 32' 

51 56' 

58 46' 


61 44' 

58 46' 

64 20' 


66 59' 

64 20' 

69 20' 


71 32' 

69 20' 

73 36' 


75 32' 

73 36' 

77 22' 


79 6' 

77 22' 

80 46' 


82 20' 

80 46' 

83 50' 


85 15' 

83 50' 

86 38' 


87 56' 

86 38' 

89 12' 

Angle at 

Af/Mo at 

Upper Limit 

Effective Center 

























"Lumen factor" for each zone is 0.9857/M 2 

equally useful with any mirror from //l.O to //3.5 in speed. The 
number of zones indicated in Table 2 has been found. sufficient to pro- 
duce results of suitable accuracy for motion picture projection sys- 
tems. The effective center of each zone is the viewing angle which 
divides the zone into two parts of equal lumen factor. 


In order to calculate the light passed by an aperture in the film 
gate, it is necessary to superimpose the aperture upon the illumina- 
tion zones in the manner as shown in Fig. 6. This figure is drawn for 
an axial mirror magnification of M 5.4, which corresponds to an 
f/2.0 mirror of about 135-degree collection angle (see Fig. 3), and for 
the sound projection aperture 0.600 X 0.825 inch. Zones 0, 1, and 2 are 
entirely within the aperture. Zones 3, 4, and 5 are partially outside, 



Vol 49, No. 3 

the percentage within . the aperture decreasing with increasing zone 
number. All zones beyond zone 5 are completely outside the aperture. 
The complications involved in determining the partial areas of 
circles superimposed upon a rectangle would add considerably to the 
difficulty of calculating light passed by the aperture for a large number 
of optical conditions of different magnification. To eliminate this, a 
short-cut method which takes into account the average distribution 
characteristics of carbon arcs has, therefore, been developed. As a 
first step, a circle of a diameter equal to nine tenths the aperture width 
is drawn, as shown in Fig. 10. In all cases with carbon arcs, the film 



i .825" X.9 =.7425 , 
V =18.85 MM./ 




d 20, 40 60 80 I007o 

FIG. 10. Relation between luminous flux within 
0.600- X 0.825-in. aperture and that within circle of 
diameter 0.9 times the long dimension of the aperture. 

gate is illuminated most brightly at the center, and from this point 
the illumination decreases progressively in all radial directions. It is 
thus possible to prepare the chart shown at the right of Fig. 10, 
relating the lumens through the rectangle to those through the circle, 
for varying rates of illumination decrease away from the center. 
This chart is an empirical one, prepared from the calculation of many 
values of lumens through the aperture over the distribution range 
indicated. Values were obtained by the more precise method of 
determining the fractional parts of the circular zones shown in Fig. 6. 
In the simplified calculation, the determination of the lumens within 
the circle is a much easier task. A similar relation can be deter- 
mined for apertures of other dimensions. 



The method for taking and evaluating brightness-distribution data, 
including further assumptions made to simplify the procedure, is 
best described by carrying through an example. The new 13.6-mm 
Super High- Intensity Positive Carbon at 290 amperes has been 
chosen for this purpose. 4 Although this carbon is not intended for 
use with mirrors, it serves as an excellent example to demonstrate the 
type of information available from this method for calculating screen 
light. Fig. 11 shows brightness-distribution curves along the hori- 
zontal and vertical diameters of the crater area, as viewed from vari- 
ous angles in a horizontal plane, from one side of the carbon axis. 
Six of the mirror zones listed ill Table 2 have been used, permitting 
calculation up to a collecting angle of 138.7 degrees. These bright- 
ness-distribution data show many of the features common to high- 
intensity carbon arcs of high brightness. Both the central and the 
maximum crater brightness decrease as the angle of view increases. 
The shape of the curves changes considerably with increasing angle 
of view. The effect of the foreshortening illustrated in Fig. 2D is 
apparent in the horizontal distribution curves, where a short vertical 
line has been drawn near the zero axis, at the left, to indicate the edge 
of the crater. At high values of 0, there is considerable brightness 
outside the crater (particularly in front) which is in a location con- 
tributing to light on the film gate. 

It has been demonstrated that the amount of data given in Fig. 11 
is adequate to yield results of reasonable accuracy, although a com- 
plete analysis of the crater would include views of the arc from both 
sides, as well as from areas of the mirror above and below the hori- 
zontal. Moreover, from each point of view, measurement of the 
brightness of all parts of the crater not included in the two diameters 
would be required. The following discussion justifies the simpli- 
fication indicated by Fig. 11. 

Only views from one side of the arc need to be considered because 
of two features common to the brightness distribution across the 
crater of almost all arcs. There is a vertical plane of symmetry pass- 
ing through the centers of the crater, the tail-flame and the negative 
carbon. This allows the evaluation of all side views from one side of 
the arc only. In views of the crater varied in the vertical direction, 
there is considerably more luminous flux projected above and con- 
siderably less projected below the positive carbon axis than is 
projected along it, because of the presence of the tail-flame and of 



Vol 49, No. 3 





















































1 , 








(O -H 
































n hl 






' s >- 






Ki _ 




f rv 


























* Ld 



- t 




















f ) 

































60 CE 

































o 1 


































C ^3 


o T 


C/l CQ 

cvj co 

8 8 

(\J 00 ^ 

- SS3NlH9iaa 


the negative carbon. However, the average of the radiations at any 
given viewing angle above and below the axis is near enough to that 
at the corresponding angle in the horizontal plane so that the hori- 
zontal measurements alone are sufficient. As far as the mirror itself 
is concerned, limitation of the measurements to angles of view in a 
single plane to one side of the arc is justified because the mirror is a 
surface of revolution about the positive carbon axis. The dimensions 
of the crater image formed on the film gate are a function only of the 
viewing angle 0, irrespective of whether the view is up, down, or from 
the side. Thus, brightness-distribution measurements made at a 
viewing angle 6 to the side can be assumed to apply to the complete 
annular zone of the mirror for which 6 is constant. 

Limitation of the brightness measurements to the horizontal and 
vertical diametral lines on the crater area has been justified by a 
comparison between crater candle-power values calculated from the 
horizontal and vertical data alone, and those calculated from a com- 
plete integration of the brightness over the entire area of the crater. 
Some cases have been found, particularly at large values of 6, where 
considerable disparity exists between candle power determined in 
these two ways. The brightness-distribution method can be made 
to agree with the precise integrating method by including brightness 
data from two 45-degree traverses of the crater. 

The data in Fig. 11 are plotted with no dimensional change in the 
crater except that which results from the foreshortening. The varia- 
tion in mirror magnification with angle' of view is taken into account 
at the next step in the procedure. A set of transparent scales has 
been made, one for each equal-lumen mirror zone, for tabulating the 
brightness-distribution data at the radii which will be centered in the 
0.5-mm image zones on the film gate. To make the distances on the 
crater as plotted in Fig. 11 correspond to the image distance at the 
gate, each scale is drawn with the spacing between divisions expanded 
as required by the zonal-magnification ratios listed in Table 2. The 
scales may then be laid over the corresponding brightness-distribution 
curves, and the brightness values at the intersections with the scale 
lines tabulated for the appropriate zones. 

From each angle of view of the crater, four such brightness values 
are thus tabulated for each film-gate zone, two from the horizontal 
brightness-distribution curve, and two from the vertical. These are 
grouped with the corresponding values from other angles of view to 
obtain the average brightness of each film-gate zone. The averaging 

238 JONES Vol 49, No. 3 

at this stage introduces another assumption. The same symmetry 
which permitted the limitation of brightness measurements to angles 
of view in the horizontal plane to one side of the arc only, allows the 
brightness effective for any crater zone to be represented with suffi- 
cient accuracy by the average of four readings in the zone, two in the 
horizontal and two in the vertical diametral lines of the crater. 

As an example, such a calculation will be carried through for film- 
gate zone 5. For unit magnification, the center of this zone is lo- 
cated 2.5 mm from the center of the film gate. The location of zone 
5 as indicated by the transparent scales on each curve in Fig. 11 is 
shown by the vertical dashed lines. The brightness values so indi- 
cated are tabulated in Table 3, together with the average brightness 
for each mirror zone. The luminous flux from each mirror zone listed 
in Table 3 is obtained by multiplying the average brightness by the 
lumen factor (0.9857) and by the area of the film-gate zone (7.85 mm 2 
for zone 5). 


Average Brightness and Luminous Flux on Film-Gate Zone 5 for New 13.6-Mm 
Super-High-Intensity Positive at 290 Amperes 


Brightness C/Mm 2 on Diametral Line 
Horizontal Vertical 

for Mirror 

Lumens from 
Each Mirror 
























* 862 



















The effective brightness and the total lumens projected on this 
zone by mirrors of collecting angles defined by each mirror zone in 
turn are listed in Table 4. 


Effective Brightness and Total Lumens on Film-Gate Zone 5 from New 13.6-Mm 
Super-High-Intensity Positive at 290 Amperes 

Mirror- Effective 

Collecting Brightness, Lumens on 

. Angle, Degrees C/Mm 2 Zone 5 

62.7 1174 9084 

86.6 1066 16489 

103.9 1002 23252 

117.5 961 29736 

128.7 909 35160 

138.7 865 40100 


Here, the effective brightness is the cumulative average of all the 
brightness values tabulated for mirror zones within the corresponding 
collection angle. Similarly, the lumens are the cumulative total of 
the lumens from each included zone. 

To complete the picture at the film gate, this same procedure is 
carried through for each film-gate zone within the area of interest. 

For greater usefulness, the luminous-flux and effective-brightness 
values may be plotted as shown for two collecting angles in Fig. 12. 


xlO 3 

1 00 




































^ N 


' / 






) 46802468 


FIG. 12. Distribution of luminous flux on film gate when M = 1 for 
mirrors of collecting angles 103.9 and 138.7 degrees with new 13.6-mm 
super-high-intensity positive carbon at 290 amperes. 

Here, in the upper pair of curves, the lumens are added cumulatively 
from the center of the film gate, so that the total lumens within any 
circle of interest can be conveniently determined. The ratio of in- 
tensity at the edge of the circle to that at the center can be read di- 
rectly from the bottom curves. Here the ordinates were obtained 
by dividing the effective brightness of the outer film-gate zone by 
that at the center. 

The luminous flux passed by an aperture in the film gate now may 
be calculated for any desired mirror, in the following manner. An 
//2.0 mirror of 103.9-degree collecting angle will be taken as an ex- 
ample. Reference to the mirror characteristic curves of Fig. 3 shows 

240 JONES 

that such a mirror has an axial magnification of 3.95. Since the calcu- 
lations have been based upon unit axial magnification to give zones 
of 0.5 mm width at the film gate, this mirror, with axial magnification 
3.95, will collect and distribute the same luminous flux per crater 
zone into zones of width 3.95 times 0.5 mm at the film gate. To find 
the luminous flux within a given circle upon the film gate, the ab- 
scissas in Fig. 12 may be expanded by a factor of 3.95 to fulfill this 

The luminous flux through the 35-mm sound projection aperture is 
desired in this example. Accordingly, from Fig. 10, the radius of the 
circle for side distribution position is 18.85/2 mm (distance from C 
to S). Dividing by 3.95 yields 2.39 mm as the equivalent radius for 
use with the unit-magnification graph. On Fig. 12, the side-to- 
center distribution on the aperture for the 103.9-degree collecting 
angle is 76 per cent at 2.39 mm. The luminous flux within a circle 
of 2.39 mm radius is 64,000 lumens. As shown in Fig. 10, 76 
per cent distribution ratio is associated with a rectangle-to-circle 
lumen ratio of 1.103. Therefore, an //2.0 mirror of 103.9 degrees 
collecting angle will deliver 64,000 times 1.103, or 70,000 lumens 
through the aperture, with the carbon arc in question. 


1 JONES, M. T., ZAVESKY, R. J., AND LOZIER, W. W., "Method for Measurement 
of Brightness of Carbon Arcs," /. Soc. Mot. Pict. Eng., 45 (1945), pp. 10-15. 

2 BENFORD, FRANK, "Studies in the Projection of Light," Series of 22 articles in 
Gen. Elec. Rev., 26-29 (1923-1926). 

3 BENFORD, FRANK, "The Projection of Light," J. Opt. Soc. Amer., 35 (1945), 
pp. 149-156. 

4 JONES, M. T., ZAVESKY, R. J., AND LOZIER, W. W., "A New Carbon for In- 
creased Light in Studio and Theater Projection," J. Soc. Mot. Pict. Eng., 45 
(1945), pp. 449-458. 





Summary. Equipment designed to accomplish eardrum and other macrophotog- 
raphy is described as well as the construction and functional operation of equipment 
for photomicrography. 


The Type of Work Done. Our intention is to show how 16 mm 
motion picture equipment has been adapted for medical and scien- 
tific use. We should like first, however, to describe the type of work 
which is done so that you may more easily see how valuable and 
necessary these adaptations are. The work done is the planning 
and production of medical motion pictures and illustration visual 
aids to learning for the medical and scientific fields. It can be 
divided into three general classifications undergraduate teaching, 
postgraduate or extension instruction, and the recording of visual 

In undergraduate teaching, subjects are prepared at class level 
and in suitable lengths for classroom periods. Films are made so as 
to be universally acceptable to all schools and to supplement the 
standard texts closely. 

The majority of over 400 films on the approved list of the Ameri- 
can College of Surgeons were designed for postgraduate and exten- 
sion instruction. A physician graduating 5 or 10 years ago is 5 or 
10 years behind in his profession unless he has been able to keep up 
to date by selected reading and regular visits to the teaching clinics. 
If he were able to read all the acceptable material printed and added 
to the medical libraries each year, he would have to read several 
volumes a day. If he did this, obviously he would have little time 
to practice his profession. Accordingly, motion picture digests of 
selected subjects dealing with new and approved techniques and 
treatments are prepared. These are approved by authoritative 
groups, and made available to the practicing physician through 
State and County medical societies, hospital staffs, and allied groups. 

The recording field covers many applications where visual 
records are desired for careful analysis and subsequent study, such 

* Presented Apr. 21, 1947, at the SMPE Convention in Chicago. 
** Medical Motion Picture and Illustration, Chicago, 111. 
f Bell and Howell Co., Chicago, 111. 


242 LA RUE AND LA RUE Vol 49, No. 3 

as the recording of clinical findings for later study and comparison 
or the recording of observed phenomena for repetition at will. 

The work is performed under the direction of such groups as the 
American Association of Medical Colleges, American Medical Asso- 
ciation, and the American College of Surgeons. Photographic 
material is secured at individual medical schools, teaching hospitals 
and clinics, and research laboratories. The work is financed by 
established Foundations and by grants-in-aid from research groups 
and ethical pharmaceutical manufacturers. 

The Need for Special Apparatus. Much of the work, such as the 
photographic recording of clinical and surgical procedures, labora- 
tory techniques, and similar items, can be adequately handled by 
the conventional 16-mm motion picture camera with a full comple- 
ment of lenses up to 6 in. in focal length. Most lenses, however, 
will not focus at distances less than l l / 2 or 2 ft. This leaves a vast 
no-man's land which can only be handled by special equipment or 
adaptations of existing equipment which will permit the making of 
macroscopic or microscopic motion pictures. 

Paul Holinger has pioneered in the development of motion and 
still cameras for various 'types of endoscopic photography. His en- 
thusiasm and unselfish assistance has made possible the development 
of similar equipment for macroscopic and other related purposes. 
Such a unit was described by La Rue, Sr., and Brubaker in the Sept., 
1946, issue of the Journal of the Biological Photographic Association. 

The Macroscopic Camera. This unit, built around the Filmo 
Auto Load 16-mm cartridge camera, (Fig. 1), records minute areas 
and objects down to the point where they may be photographed 
under the low-power microscope. The mechanical arrangement of 
the unit provides for the support of light source, condenser lenses, 
water cell, 45-deg mirror, the Auto Load camera, and focusing tele- 
scope. A focusing knob with rack and pinion moves the whole ball- 
bearing-mounted assembly to focus on the area desired. An ad- 
justable friction mechanism permits smooth operation and the 
locking of the camera at any desired focus. 

The image on the film may be adjusted from one-half life-size to 
full life-size. At one to one, or actual size, it is possible to resolve 
clearly objects as small as 0.002 in. The very fine hairs found 
within the entrance to the ear canal can be seen well defined. The 
films of the eardrum, for which the instrument was primarily de- 
signed, clearly show the radiate fibrous layer of the drum membrane, 

Sept. 1947 



the radial lines being very prominent. There are, of course, many 
other applications for such a unit. 

The light source is a 6-v 18-amp ribbon-filament lamp (Fig. 2). 
The 2-mm wide tungsten ribbon filament is focused by means of the 
condenser lenses upon the object plane. A 5-mm wide image of the 
filament is produced. The lamphouse is movable in relation to the 

FIG. 1. General view, showing base, rods, bearings, and lens blocks. 

A, Lamp base; B, Water cell; C, Mirror; D, Lens; E, Beam-splitter housing; 
F, Camera; G, Focusing telescope; H-H, Ball-bearing rollers on rod; /-/, Adjust- 
ment locks to control magnification factor; /-/, Condenser lens. 

condenser lenses and thus provides for a range of adjustment of the 
image at the object plane. For surface macrocinematography, the 
filament image is projected in the same plane on which the camera 
is focused. At the lower-magnification ratios the image is thrown 
out of focus to illuminate a wider band of the object area. 

A beam-splitter cube, immediately behind the camera lens, 
diverts a portion of the light to the erecting system and eyepiece of 
the focusing telescope. The remaining light forms the image on 
the photographic film. 

The camera-release button is controlled by a solenoid. A beveled 
ring attached to a snap-action switch lever is located behind the 
focusing knob so that the camera can be focused, locked, started, 
and stopped with the right hand without removing the hand from 



Vol 49, No. 3 

the knob. This control ring operates the switch and only a slight 
pressure of the thumb is required to operate it. 

With constant supervision of focus by means of the telescope, 
instantaneous decisions can be made as to starting and stopping the 
camera or for corrections in focus while the camera is in operation. 

Microcinematography in 
Color. Much of the progress 
of med'ical science has been de- 
pendent upon microscopy. 
Many disease entities and most 
pathology can only be recog- 
nized under the microscope. 
Photomicrography (the photo- 
graphing of specimens under 
the microscope) has been an 
important method of illustrating 
scientific papers and classroom 
lectures for a number of years. 
Most medical colleges have 
photographic departments 
equipped to make photomicro- 
graphs of selected specimens 
and fields as a matter of routine. 
The manufacturers of micro- 
scopes and microscopic equip- 
ment have manufactured equip- 
ment for this purpose. 

With the advent of color, 
however, serious problems arose. 
All available color films were 
greatly reduced in speed and the 
maintenance of color temperature became an important factor. The 
making of black-and-white motion pictures through the microscope 
has always been a difficult procedure, but the making of color films 
at normal speed seemed almost impossible. Several microscope 
and camera manufacturers suggested several methods of accomplish- 
ing this, but-none of the methods investigated completely answered 
the purpose. In order to meet the many requirements in micro- 
cinematography, it became necessary to develop an apparatus 
which would permit the making of motion pictures of objects under 

FIG. 2. Rear view showing focusing 
target (removable when working, only 
to be used to check focus and align 
lamp), focusing knob, microswitch, 
focusing telescope, and other parts. 

A, Beam-splitter cube; B, Focusing 
telescope; C, Mirror; D, Focusing tar- 
get (removable) for aligning lamp and 
checking focus; E-E, Condenser lens; 
F, Water cell; G-G, Light source; H, 
Release ring; I, Focusing knob. 


the microscope while being viewed by the operator. An apparatus to 
permit this is pictured in Fig. 3. 

In most cases it is necessary to go where the work is being done 
so that any equipment designed for this purpose had to be extremely 
light and portable. It was desirable to employ the standard type 

FIG. 3. General view of microscopic setup. 

A, Compensating eyepiece; 'B, Lamphouse for ribbon filament lamp and No. 1 
photoflood; C, Light trap; D, Reflex focuser; E, Camera (without lens); F-F, 
Adjustment for height; G-G, Hollow steel rods; H-H, Magnesium channels; 
/, Sponge-rubber pads; J-J, Lock position; K'-K, Microscope held in jig; L, 
Filter holder ; M, Water cell ; N, Matched compensating eyepiece in light trap ; 
0, Microstage; P, Beam-splitter cube; housing replaces the nosepiece of the 

of microscope and attachments, and it was essential that the opera- 
tor be able to view the field at the time the pictures were being 
made. For practical and economical reasons, the apparatus was 
designed to accommodate conventional 16-mm motion picture 
cameras. It is needless to say that the unit has to be sturdy and free 
of vibration and that the "setup" be simple and speedy. So that it 
might be operated in a normally illuminated room, it was necessary 



Vol 49, No. 3 

to have a light-tight optical path from the microscope to camera. It 
was found that such an apparatus could be constructed with the 
optical system shown in Fig. 4. 

Although this principle is well known, this particular application 
is the result of the efforts of J. D. Brubaker and W. B. Park. At the 
left is shown the conventional microscope optics. A split-beam 
cube reflects the greater portion of the image beam to the eyepiece 
just before the film plane. Two matched eyepieces are employed. 
Though it is possible to obtain an image on the film without the use 
of an eyepiece near the film plane, the employment of this eyepiece 
corrects for objective aberrations and further simplifies operation. 







FIG. 4. Optical diagram. 

In Fig. 3 is seen the result. The base is constructed of magnesium 
channels and the optical rods, or bench rods, are of hollow steel and 
mounted in adjustable blocks for initial alignment. The microscope 
hold-down blocks are of bakelite and a magnesium rod extends from 
the back with threaded holes to support the various types of light 
sources used. This rod is removable to simplify transportation from 
place to place. A conventional type of microscope is employed, the 
nosepiece being replaced with the split-beam housing. The micro- 
scope objectives and eyepieces are standard arid the stage of the 
microscope is one of the better micro types. The camera support is 
adjustable in height to compensate for the various objectives used. 

On location it is a simple matter to set up the apparatus in a con- 
venient place for operation. This is usually on a table or desk in or 
adjacent to the clinic or laboratory. The microscope is fixed in its 


jig and locked into position. The extension tube from the split-beam 
housing is screwed into position and the light trap attached. The 
camera is then mounted on its tripod screw and a Goerz reflex focus- 
ing device is screwed into position taking the place of the lens. On the 
end of the reflex focusing device the other part of the light trap is 
screwed into position . The light source is then mounted and centered 
in the usual manner. The compensating eyepieces are matched as 
to f oc*us the camera being moved along the rods to the extreme right 
position so as to permit inspection through the camera eyepiece. 

When the whole apparatus is set up and adjusted, it is mounted 
on sponge-rubber pads to prevent vibration from being transmitted 
to the apparatus. While there is some vibration during the course 
of exposure, the whole apparatus vibrates as a unit and motion 
pictures have been made at speeds up to 64 frames per sec without 
any unsteadiness on the screen. The light source employed is a 6-v 
18-amp ribbon-filament lamp for which a photoflood may be substi- 
tuted when using a 25-mm or longer focal -length objective. The 
lamphouse is especially designed to accommodate both the photo- 
flood and the ribbon-filament lamp together with the standard water 
cell and filter holders. 

The use of the reflex focuser increases the magnification to some 
extent because of the lengthened optical path. Its use is desirable 
because it permits a constant check on the image reaching the aper- 
ture but is not used if magnifications must be exact. When the mag- 
nification must be exact with that viewed through the eyepiece, the 
reflex focuser is dispensed with and focusing is accomplished with a 
prism in the aperture of the camera. 

All the possibilities surrounding the use of this equipment have 
not been explored. As an example, Nicholl and Webb of the Uni- 
versity of Indiana have been engaged for the past several years in 
blood-circulation studies utilizing the wing of the bat. They have in- 
spected and tested this equipment and as a result a similar apparatus, 
with a modification to permit the use of a dissecting microscope, has 
been designed. Their desire is to record their findings visually dur- 
ing their research so as to allow repetition and group study at leisure. 

Animation. In the production of medical teaching films, much 
of the information and material which cannot be pictured by con- 
ventional methods can be diagrammatically portrayed by means of 
animation on 16-mm color film. This also demands special equip- 
ment. The subject, however, is beyond the scope of the present paper. 



Summary. Special photographic equipment and techniques have been developed 
for motion picture photography of the human air and food passages. These films 
graphically visualize the vocal cords, windpipe and bronchial tubes, and the esophagus 
from the mouth to the stomach, to provide unusual clinical records that are invaluable 
as teaching and research material. 

The camera developed for this work permits constant visualization through the 
bronchoscope for finding and focusing as well as during the actual filming. 

Advances in the field of photography have left few of the body 
cavities inaccessible to the camera. Such photography is an impor- 
tant method of recording the normal anatomy and diseased states of 
these areas. This paper concerns photography of the cavities visu- 
alized through the mouth, including the interior of the mouth and 
nose, the vocal cords, the windpipe, the bronchial tubes and the esoph- 
agus; these constitute the respiratory tract and the food passage from 
the mouth to the stomach. Films made of these areas depict the 
normal respiratory and swallowing functions, and assist in the study 
of the action of the vocal cords and the mechanism of speech produc- 
tion. As clinical records, the films permit careful study of tumors, 
inflammatory processes, and even the mechanics of diagnosis and 
manipulation for removal of bizarre objects such as pennies, safety 
pins, tacks, and similar items that find their way into the air and food 
passages of infants and children. 

Before proceeding with a description of the camera, an analysis of 
the problems encountered should be described. The vocal cords may 
be photographed through a tube approximately 5 /s to 3 /4 in. in diam- 
eter and 6 to 8 in. in length. The tube for photography of the bronchi 
a'nd esophagus may have a maximum diameter of 5 /s in., and must be 
at least 14 in. long. Only open tubes, rather than tubes containing 
lenses, may be used because it is necessary for the patient to continue 
breathing through the tube as the airway is being examined. Thus 
the problem consists in designing an apparatus for photography of the 

* Presented Apr. 21, 1947, at the SMPE Convention in Chicago. 
** Department of Otolaryngology, University of Illinois, College of Medicine, 
Chicago, 111. 

t Design and Research Engineer, 805 Greenleaf St., Evanston, 111. 



area seen through a 5 /s- or 3 /Vm. tube, 8 in. in length, and a second 
tube 5 /s in. maximum diameter, 14 in. in length. To solve this prob- 
lem it was necessary to develop special equipment which was evolved 
through a compromise of many limiting factors, some dependent 
upon fundamental photographic principles, and others dependent 
upon the configuration of the air and food passages. In designing the 
equipment, the safety of its use in the patient was considered para- 
mount. Ease of manipulation, constant visualization of the field, 
both during the introduction of the instrument and during the actual 
photography, axial illumination, and a relatively great depth of field 
were all considered to be essential features. 

FIG. 1. Left side of camera assembly. The parts, from 
left to right, are the lamp housing, attaching clamp, glass 
slide, supplementary lens slot, focusing knob, masks, and 
the camera box. Above the lens slot and focusing knob 
are the focus indicator, telescope housing, and eyepiece. 


The new endoscopic motion picture camera is similar in some re- 
spects to a type previously constructed. *~ 4 The light-source, focus- 
ing telescope, and camera are combined in one unit (Figs. 1 and 2). 
An attaching clamp at the front of the unit permits the instant attach- 
ment or detachment of any of several endoscopes or light-reflecting 
tubes, used to photograph the various areas. A heated removable 
glass slide placed immediately behind the attaching clamp shields the 



Vol 49, No. 3 

optical parts from gross soiling or condensed moisture. The appa- 
ratus is held by a built-in handle when the operator is photographing 
through the endoscopes, or it is supported on a tripod for photography 
of the larynx by indirect mirror method. The camera is started and 
stopped by a trigger lever in the handle which also automatically 
raises the lamp voltage to the proper color temperature while the 
camera is running. The plane of sharp focus at the object is adjust- 
able and may be placed at any position from 9 to 26 J /2 in. from the 

FIG. 2. Interior of camera with cover and lamphouse removed. 

attaching clamp. The telescope and camera are focused simultane- 
ously by a knob on the left side of the housing. The telescope shows 
an exact duplicate of the film image during finding and focusing as 
well as while the camera is running. A slot permits the insertion of a 
supplementary achromatic lens of 24.4 cm (9 5 /s in.) focal length for 
photography of the eardrum at a relatively high magnification. 

The endoscopes and light-reflecting tubes are polished and nickel- 
plated on their inner surfaces. Near the tip of each endoscope the 
interior surface is threaded for x /2 in. to help outline the circular field. 
This is necessary since it is impracticable to arrange the masking 


device to mask close to the circular film image. Thus, if the endo- 
scopes are bent slightly during filming, the image is allowed to change 
position in relation to the mask opening without wandering off the 

The attaching clamp permits instant attachment or detachment of 
any of the several endoscopes or light-reflecting tubes. The clamp 
makes a solid mechanical connection between the endoscope and the 
camera. The endoscope axis is aligned accurately with the camera 
axis so that the image of the endoscope tip comes in the center of the 
camera film aperture, and is approximately centered in the mask 

A heated glass slide is used immediately behind the attaching clamp 
to shield the optical parts of the camera from gross soiling or con- 
densed moisture. A rectangular opening at one side of the front face 
of the glass slide allows passage of air through the endoscope to per- 
mit the patient to breathe freely when the bronchoscope or direct 
laryngoscope is used. The glass is optically flat and introduces no 
aberration into the film image. It is placed at an angle of 5 deg to 
avoid reflections. After the initial heating of the slide, the heat 
absorbed continuously from the camera lamp bulb keeps the slide 
warm enough to avoid fogging. 

Illuminating System. An airplane headlight bulb is used as the 
light source in this camera. It was chosen because its two compact 
filament coils occupy a very small space about 5 mm square, and be- 
cause it can be operated on its side instead of in an upright position. 
The technical description of this airplane tungsten filament lamp is 
240 w, 12 v, 20 amp; medium prefocus base; C-2 type filament; 
A-19 bulb (2 3 /s in. in diameter) ; total lamp length 4 x /8 in. 

The light from this lamp is directed onto the main axis of the cam- 
era by means of two condenser lenses and the plane mirror at 45 deg 
(Fig. 3) . A spherical mirror behind the lamp increases the light level 
and produces more even illumination. An enlarged image of the lamp 
filament is projected to a plane about 5 in. ahead of the attaching 
clamp. This is the optimum distance in order to produce the maxi- 
mum light level through the endoscopes and light-reflecting tubes. 
A heat-absorbing glass is used between the condenser lenses to absorb 
excess heat from the lamp. This is necessary because the heat reach- 
ing the tip of the endoscope is slightly greater than the mucosal sur- 
faces can tolerate with safety. 

The lamp is operated at 12 v while finding and focusing, and is 



Vol 49, No. 3 

fHfg:i:!iiivr' a on 

'= I s *sj= P.I_"| "5^ iasjtifll 


raised automatically to 14y 2 v while the pictures are being taken, by a 
relay operated by the camera trigger. This produces light of approxi- 
mately 3450 K, the correct color temperature for type A Kodachrome 
film. The estimated life of the airplane headlight lamp at 12 v is 
about 50 to 100 hr. When operated at 14 x /2 v, the life is about 30 
min actual filming time, which permits approximately 750 ft (15 
magazines) of film to be taken with one lamp. 

Since the endoscopes are polished and nickel-plated on the inside, 
photographic light intensity of about 160 foot-candles is obtained at 
the endoscope tip when the lamp is operating at 14V2 v. This cor- 
responds to a lens setting of f/S at x /4o sec for type A Kodachrome 
film. Since light is absorbed in the glass slide and beam-splitter cube 
the true stop is about //6.8 at the camera lens when average-colored 
areas are photographed. When the supplementary lens is used for 
eardrum photography, the light intensity at the eardrum, through the 
ear speculum, is about 650 foot-candles, corresponding to a lens setting 
of //1 6 when a lens is focused at infinity. The supplementary lens 
produces an image-to-object ratio of 0.50, and although some light is 
absorbed in the beam-splitter cube, a marked /stop setting of //ll to 
//1 6 can be used. The light distribution at the tip of the endoscope 
and at the tip of the ear speculum is even, and no image of the lamp 
filament on the field is noticeable in the finished pictures. 

In order to check the actual photographic light intensity at the tip 
of the endoscopes and light-reflecting tubes, a General Electric ex- 
posure meter is used. This permits a periodic over-all check of the 
proper lamp alignment, lamp blackening, and internal endoscope re- 
flectivity ; these factors may affect the proper exposure of the film by 
causing a reduced light intensity. The meter is used in an empirical 
manner. The incident-light method is used, with a 1 /g-in. hole in an 
opaque mask covering the photocell. A standard light value is de- 
termined by measuring the light 'intensity at the endoscope tip which 
produces satisfactorily exposed .film, with proper color rendition. A 
reduction of more than 25 per cent in light value usually requires re- 
placement of the lamp bulb. The light intensity beyond the endo- 
scope tip is uniform enough for satisfactory photography of areas up 
to about 3 or 4 in. beyond the endoscope tip. For example, a normal 
level of illumination is obtained at the plane of the larynx when the 
mirror tube is used. In this case, the photographic field is more than 
3 in. from the tip of the light-reflecting tube. 

The electrical circuits reaching the camera are at a low voltage and 



Yol 49, Xo. 3 

are well insulated from the 115-v a-c line (Fig. 4). The low- voltage 
circuits are insulated from the metallic camera housing, and the trans- 
former windings supplying the low voltages are well insulated from 
the line. With this arrangement, complete safety is provided against 
electrical shocks to operator or patient. 

Optical System. The basic arrangement allows both the camera 
lens and the focusing telescope to operate on the same axis, and use 
the same supplementary lens for higher magnifications. Figs. 2 and 
3 illustrate the following discussion. 

210 WATT, 1 2 VOL T 

FIG. 4. Electrical and control diagram of camera. 

Through the use of a "beam-splitter cube", the camera and focus- 
ing telescope view the field at the same time and on the same axis. 
This beam-splitter cube consists of two right-angle prisms with their 
hypotenuses cemented together. Before cementing, the hypotenuse 
of one prism is coated with a very thin partially reflecting 
aluminum coating. Most of the light reaching the cube passes 
through it to the camera lens. About 15 per cent of the light is di- 
verted at 90 deg to the telescope objective lens. The camera lens 
and telescope-objective lens are identical. They are a matched pair 
of 90-mm focal length Wollensak Velostigmats, Series 2. The lenses 
are of exactly equal focal length, and are of //4.5 aperture. They are 
of the coated type (with antireflection coating) for improved light 
transmission and improved image contrast. The maximum usable 


aperture for both lenses is limited to //5.6 by the dimensions of the 
beam-splitter cube, and by the tube through the 45-deg mirror. 

The two lenses are moved simultaneously by a knob with rack-and- 
pinion motion, through equal distances, and without any backlash or 
play in the gearing. Friction holds the focusing knob wherever it is 
stopped. The distance from each lens to its own focal plane is always 
identical with that of the other lens. Since the focal planes are fixed 
in position, the motion of the lenses moves the object plane of sharp 
focus farther or closer in relation to the attaching clamp, and in this 
way, the camera lens and telescope are always focused upon the same 
plane in the object space. The supplementary lens simply shifts 
both planes of sharp focus a certain distance toward the attaching 
clamp. With the supplementary lens, the focusing action is basically 
the same as with the regular endoscopes. The regular focusing range 
of adjustment is from 9 to 26V2 in. from the attaching flange. At the 
supplementary distance, the useful range is from 4 to 5 in. from the 

The focusing telescope is an extremely important part of the 
camera. By means of the telescope, the field at the endoscope tip is 
constantly under visual supervision, allowing safe manipulation of the 
endoscope in order to reach the desired field for photography, or for 
supervision of the field while the endoscope is advanced or withdrawn 
while filming. The telescope enables the plane of sharp focus for the 
camera to be placed at any distance beyond the endoscope tip ; the 
plane of sharp focus may be changed while the camera is running. 
The field is seen at all times during filming as well as before and after 
the pictures are taken. Experience has shown that the mucosal sur- 
faces of low contrast can be focused upon more accurately by using the 
aerial image and therefore no ground glass is used at the eyepiece 
focal plane. The telescope is adjusted to produce correct focus for 
the camera when the eye is relaxed and focused upon infinity similar 
to looking into a microscope. The eyepoint is far enough away from 
the eyecap to permit the operator to wear glasses. The camera is de- 
signed for a right-handed person, and for the right eye at the telescope. 

The image in the telescope eyepiece is erect and correct from side to 
side, the same as though the operator were looking directly into an 
endoscope. The optical system must provide this correct image, 
since an inverted or transposed image introduces serious risk of 
trauma when the endoscope is manipulated. The endoscopist would 
attempt to move the endoscope tip in the opposite direction to that 


desired. In the present camera, the erect image is accomplished by 
means of an erecting lens, rather than a porro system of prisms. The 
three prisms 5 shown in Fig. 3 are used as front-surface mirrors; this 
is done to bend the telescope axis in order to place the eyepiece in the 
most convenient position. The image seen in the telescope is an exact 
duplicate of the picture recorded on the film. The total magnifica- 
tion of the field as seen in the telescope eyepiece is about five times for 
the endoscopes, and about ten times for the supplementary lens dis- 
tance for eardrum photography. 

A focus indicator is provided in a window at the side of the main 
housing to show the position of the plane of sharp focus ahead of the 
attaching clamp. The indicator may be moved by the focusing knob 
to the position marked on the scale corresponding to the endoscope 
used. Without such an indicator, it would be necessary to move the 
focusing knob while observing through the telescope with a sterile 
cloth held just beyond the endoscope tip in order to make the focus set- 
ting a source of annoyance and lost time in the busy operating room. 
The presetting method by means of the indicator is sure and rapid. 

Camera Film Box and Photographic Lens. The motion picture 
camera box is a Bell and Howell Auto Load. It is permanently 
attached to the main housing of the endoscopic camera. The release 
button is operated by the trigger in the handle. The Auto Load 
camera uses 16-mm type A Kodachrome film in 50-foot pre threaded 
magazines and runs about 12 1 /2 ft at one winding. This corresponds 
to 31 sec of filming time at 16 frames per sec. While the normal shut- 
ter speed of the camera is l /& sec at 16 frames per sec, the shutter- 
blade opening on this camera was increased to allow the light from the 
lens to remain on the film for a longer period of time, permitting the 
use of a slightly smaller / stop. The new shutter speed is Vs4 sec. 
This change is possible because the film image is limited to a circle 
0.250 in. in diameter. The camera model chosen provides for the 
choice of 8, 16, 24, or 32 frames per sec. A speed of 16 frames per sec 
is ordinarily used, but a speed of 24 frames per sec is possible at//6.3 
for photography of the larynx where much light is reflected from the 
photographic field. 

A special masking device is built into the left side of the camera 
film box in order to mask off the out-of -focus images reflected from 
the inside of the endoscopes (Fig. 3) . The mask is a metal strip with 
a series of four holes of graduated sizes corresponding to the film- 
image diameter of the various endoscope tips. The largest opening is 


0.250 in. in diameter and the strip is moved by a small knob (Fig. 1) 
to center each hole automatically in the film aperture. The mask is 
0.250 in. ahead of the film, and it masks fairly close to the circular 
film image, with only a small margin of the endoscope wall visible in 
the finished picture. Without such a mask, the bright out-of -focus 
images reflected from the inside of the endoscopes would cause con- 
fusion when the finished film is viewed. 

As previously stated, the camera lens is a 90-mm focal length //4.5 
anastigmat type. The maximum usable aperture is //5.6. The / 
stop is adjusted by a knob at the rear of the main housing. / stops of 
5.6, 6.8, 8, 9.5, 11, 13.6, and 16 are provided (one-half stop intervals) . 
The maximum circle of confusion on the film, due to aberrations of the 
lens, is less than 0.001 in. In other words, a geometric point at the 
object plane of sharp focus is imaged as a circle smaller than 0.001 in. 
at the film plane. This is entirely adequate for the best quality mo- 
tion picture photography on 16-mm Kodachrome film, and is the 
accepted standard for the best lenses used with 16-mm cameras. 

The depth of field (sometimes called depth of focus) at the object 
plane of sharp focus, is about 0.250 in. for a reasonably sharp image. 
This is for a maximum circle of confusion on the film of 0.004 in. For 
a good practical value for sharpest possible image, the depth is only 
about 0.125 in. when 0.002 is taken as the circle of confusion. 
Theoretically it is possible to show details which are imaged as small 
as 0.001 in. at the film plane, but 0.002 in. is a good working value for 
best results. Because of the motion of the screen image, occasional 
short scenes very slightly out of focus detract little from the general 

The camera handle is notched to provide a comfortable and secure 
grip. The index finger is free to operate the trigger camera release. 
The handle is located so that the center of gravity of the camera is 
balanced over the hand when the camera axis is tipped forward to 25 
deg from the horizontal, the approximate angle of the endoscope when 
the patient is in the recumbent position for direct per oral endoscopy. 
A threaded hole for a tripod is located in the bottom of the handle to 
support the camera while mirror pictures of the larynx are being 
taken, or when pictures are being taken at the supplementary distance. 


The endoscopes are heated to slightly above body temperature by 
an electric heating pad or by a brief immersion in the sterilizer just 


before use. Condensed moisture on the inner surface of the endo- 
scopes would reduce the reflection of photographic light from the 
lamp. The glass slide for the camera is heated in the lamp housing 
and the heat from the bulb keeps it warm after it is inserted in the 
camera. A spare glass slide is available should the slide in the camera 
become spattered if the patient coughs. An assistant receives the 
camera from the endoscopist when it is detached from the endoscope 
to permit cleaning, winding, or changing film. The endoscopes may 
be introduced into the patient with the camera attached, since the 
lamp and telescope give full visualization of the field at the endoscope 
tip. The quick-acting attaching clamp for the endoscopes provides 
instantaneous release of the camera for aspiration of secretions by 
suction tube if necessary. In order to observe the endoscopic field 
when the camera is detached, a light carrier from a standard broncho- 
scope of proper length is used, held inside the endoscope. Light- 
carrier canals and suction tubes are not built into the endoscopes 
since they would encroach upon the circular field and detract from the 
appearance of the finished film. 

While the endoscope is being introduced, the plane of sharp focus 
is set just beyond the endoscope tip. The moving field is photo- 
graphed to demonstrate the landmarks seen during the introduction 
of the various instruments. The plane of sharp focus is then set at a 
suitable distance ahead of the endoscope tip, and the endoscope 
slowly advanced or withdrawn with the camera running, in order to 
show the moving field as seen during an endoscopic examination. In 
this way, it is possible to take a series of scenes fully simulating a 
bronchoscopic examination. 

Contrary to general opinion, it is more difficult to obtain 'satisfac- 
tory pictures of the vocal cords and larynx by the indirect mirror 
method than through the direct laryngoscope. The usual technique 
is to support the camera on a tripod with the light-directing mirror 
tube attached. The laryngeal mirror may be attached to the mirror 
tube, or held independently in the operator's hand while the tongue 
is held out of the way by the patient. The mirror tube is partially 
inserted into the mouth and acts somewhat as a tongue depressor and 
prevents the tongue from rising into the photographic field. In the 
majority of patients, it is difficult to align the camera axis, mirror, 
and laryngeal axis simultaneously so that a good view is to be had of 
the larynx. An overhanging epiglottis often obscures the anterior 
commissure. An additional difficulty is the necessity of obtaining 


sharp focus of the vocal cords, arytenoids, and epiglottis, all at the 
same time. The vocal cords ordinarily are placed in sharp focus and 
the rest of the field is slightly out of focus, or the area of greatest 
interest is placed in sharp focus. The general effect is usually satis- 
factory even though some parts may be out of focus. 


Since the camera described is fundamentally a universal endoscopic 
type suitable for photographing through almost any of the common 
open-tube endoscopes or speculums, a brief discussion of its uses in 
proctology and gynecology is indicated. 

For proctology, a sigmoidoscope of 13 in. total length may be used. 
The largest diameter practicable is recommended in order to obtain a 
large field. The free working length of 7 3 /4 in. (20 cm) is satisfactory 
for most proctoscopic cinematography, and provides a reasonably 
large image on the film. A free working length of 6 in. (15 cm) is 
sufficient for much proctoscopic photography, and would provide a 
larger image on the film. In the latter case, the total endoscope 
length is Iiy 4 in. The simulated perspective effect is seen in procto- 
scopic cinematography as well as in per oral cinematography. The 
endoscope is first introduced with an obturator in place. The ob- 
turator is then withdrawn, and the field examined with a light stick 
held inside the sigmoidoscope. After the desired field is located, the 
camera is attached and the pictures are taken. The bowel wall may 
be scanned as the endoscope is being withdrawn while filming. Addi- 
tional descriptions of techniques, and photographic illustrations of 
proctoscopic views, have been published elsewhere. 3 

For photography of the cervix uteri, light-directing tubes of various 
diameters are suggested ; the maximum diameter usable on this cam- 
era is about 1 to iy 4 in. (2.5 to 3 cm) . Since the film area in the cam- 
era is limited to Y* in., the light-directing tubes must not be shorter 
than 10 in. total length in order to have the whole object field appear 
on the film. In this case, the field is planned to be about 13 to 15 
in. from the attaching flange of the light-directing tube. The tube 
diameter should be slightly smaller than the opening in the vaginal 
speculum, to avoid light reflections from the proximal portion of the 
speculum blades. The camera with its light-reflecting tube can be 
aligned externally with the axis of the speculum, and the pictures 
taken. It may be desirable to provide a short tubular sleeve attached 
to one blade of the speculum in order to hold the speculum axis in 


Vol 49, No. 3 

line with the camera axis. Without the alignment sleeve, it is possi- 
ble to scan the field slowly when the whole area to be shown cannot 
be included on one camera field at one time. Scanning may induce a 
sense of perspective when the films are viewed, especially if the cervix 
is moved slightly by abdominal palpation. 


By means of specially devised camera equipment, it has been pos- 
sible to photograph the human air and food passages. The pictures, 
in motion and color, enable the surgeon to record and study the nor- 
mal anatomy of these areas as well as their diseased states. The 
films are invaluable for teaching purposes since they show to large 
.groups of students and physicians the surgical field otherwise visible 
only to one individual. The actions of the vocal cords are shown 
and inflammatory processes and tumors of the air and food passages 
are recorded. Of particular interest is the fact that the techniques 
used to remove pennies, peanuts, and safety pins, as well as other 
objects from the windpipe and bronchial tubes of children may be 
demonstrated by motion pictures taken through the bjonchoscope as 
these objects are being removed. 


1 BRUBAKER, J. D.: "Proctoscopic Cinematography in Kodachrome," /. 
Biol Photographic Assn., 9 (Dec. 1940), p. 87. 

2 GARNER, J. M., NESSELROD, J. F., AND PEERMAN, J.: "Proctoscopic Cine- 
matography," Proc. Inst. Med., Chicago, 111., 13 (D^c. 1941), p. 387. 

3 BRUBAKER, J. D., AND HOLINGER, P. H.: "The Larynx, Bronchi and Esoph- 
agus in Kodachrome," /. Biol. Photographic Assn., 10 (Dec. 1941), p. 83. 

4 HOLINGER, P. H.: Chapter on Endoscopic Photography, "Diseases of the 
Nose, Throat and Ear," pp. 753-764, Jackson and Jackson, Editors. W. B. 
Saunders Co, Philadelphia, Penna., (1945). 


-MR. ELMS: I would like to ask whether it would be possible to photograph 
the inside walls of the stomach? 

DR. P. H, HOLINGER: Stomach photography is one of the most interesting 
of all photographic work in any of the body cavities. Mr. Brubaker will tell you 
that this camera has been designed so that we can use longer tubes in order to 
get into the stomach. We haven't done that yet, but photography of the stomach 
walls is possible in those patients who have direct openings into the stomach. 
I do not mean to digress medically, but there are many conditions that require 
that the patient be fed directly through the wall of the abdomen as the food 
passage has been obstructed by some disease. The stomach walls can be photo- 
graphed easily through such an opening. 


In Life magazine several weeks ago you will remember seeing an apparatus- 
used in animals for photography of the stomach wall through such an opening. 

What we hope to do before we finish is to go ahead with the photography you 
are discussing. 

MR. ELMS: You cannot photograph without making incisions? 

DR. HOLINGER: I won't say that we can't. Our camera has been completed 
so recently that we haven't had an opportunity to test this feature of it. I think 
it is possible. 

MR. ELMS: Is this equipment available commercially? 

DR. HOLINGER: I guess it is if you wish to have it built. I am sure that Mr. 
Brubaker could go ahead with its construction. As you see, it is a handmade 
type of apparatus but he could duplicate it. 

MR. ELMS: Do you have pictures of the removal of the safety pin? 

DR. HOLINGER: We could have shown you the removal of the safety pin but 
the time did not allow. We could have shown the various methods of closing 
the safety pin. Frequently we take the pins to the stomach and straighten them 
and then remove them. Rather than take the time, we showed various types 
of foreign bodies rather than techniques. To photograph the closing of a safety 
pin would take longer than we should wish to leave the instruments in the 

DR. E. W. KELLOGG: How far from the end of the tube is your first lens and 
what is the size of the opening of the tube, giving the field size? 

DR. HOLINGER: The size of the field was 6 / 8 in. in the smaller tube and 3 / 
of an inch in the larger tube. It is approximately 20 in. from the lens itself. 

Photography ahead of the actual tube itself is possible because the light is 
proximally located and shines down the tube rather than being distally located 
with the light at the end of the tube ; consequently, photography considerably 
ahead of the tube is possible with the apparatus about 4 to 6 in. 

DR. J. G. BRADLEY: Have you tried to use zirconium as a pin-point light 

DR. HOLINGER: Yes we have, but there isn't sufficient illumination for the 
particular purpose that we need. We tried that not only for the actual photog- 
raphy but for the surgical instruments themselves. It just doesn't give us enough. 
Maybe we didn't have it set up right. 




Summary. The application of the acoustic-impedance concept to the study of 
sound absorption has made it possible to explain and predict the performance of 
acoustical materials more completely. This paper reviews the theoretical and experi- 
mental work done on this subject in recent years and cites some applications to the 
design of commercial materials. 

In the commercial development of acoustical materials over the 
past forty or fifty years, the aim in general has been to obtain the 
highest degree of sound absorption consistent with acceptable struc- 
tural and decorative qualities and reasonable cost. The absorbing 
characteristics of commercial materials, particularly with respect to 
frequency, have been co-ordinated fairly well, either by design or 
fortunate circumstance, with the acoustical requirements of the 
spaces in which they are used. For example, in the field of noise 
abatement, it is well known that reduction of the high-frequency com- 
ponents of the average noise produces much greater relief from annoy- 
ance than the same reduction of the low frequencies. Accordingly, 
in the case of materials intended primarily for noise reduction it is 
not generally attempted to obtain as high absorption at the low fre- 
quencies as at the high frequencies. This is an economic advantage 
in that it permits the use of relatively thin materials and inexpensive 
constructions. In special situations, such as radio studios or other 
rooms in which an absorption-frequency characteristic approaching 
a flat curve is desired, more expensive constructions involving deeper 
air spaces or thicker materials are usually required. 

Until comparatively recently, knowledge of the relation of sound 
absorptivity to the physical properties of materials has been mostly 
empirical and qualitative. It has long been known in a general way 
that the absorptivity, defined as the per cent of incident sound energy 
absorbed, depends on the porosity, and, particularly at low fre- 
quencies, on the thickness of a material. It has been only within the 
last ten years or so, however, that it has been possible to predict with 
a fair degree of precision the absorbing characteristics of a material or 

* Presented Apr. 21, 1947, at the SMPE Convention in Chicago. 
** The Celotex Corporation, Chicago, 111. 


construction in terms of accurately defined and measurable physical 
properties. This has resulted principally from the application of 
the concept of acoustic impedance to the phenomenon of sound ab- 
sorption. Most of the- theoretical and experimental work underlying 
this development is due to P. M. Morse and R. H. Bolt, 1 and to 
L. L. Beranek. 2 - 3> 4 All experimental data given in this paper were 
obtained by the author. 


Specific acoustic impedance has been defined, by analogy with the 
older and more familiar electrical impedance, as the complex ratio of 
instantaneous sound pressure to instantaneous particle velocity at a 
given point in a sound wave or system of waves. When applied to 
the case of an absorbing material, the impedance of the material is 
taken as the value existing immediately at its surface.* Stated 
differently, the impedance of a material is the sound pressure required 
to produce unit velocity of air movement into its surface, and is thus 
a measure of the degree to which the material impedes the entrance 
and absorption of sound waves. Following electrical terminology, 
the impedance Z has magnitude and a phase angle, which may be re- 
solved into the acoustic resistance R representing that component of 
particle velocity which is in phase with the pressure, and the acoustic 
reactance X that component which lags or leads the pressure by plus 
or minus 90 degrees, respectively. In a single free progressive plane 
wave in air, the impedance is a pure resistance, having a constant 
value at all points equal to the density of air p multiplied by the 'wave 
velocity c. This is termed the characteristic impedance, or radiation 
resistance, of air, and has a value of approximately 41 centimeter- 
gram-second units. In any type or combination of waves other than 
a simple plane wave, the impedance in general has a value differing 

* Strictly speaking, the impedance at the surface of a material is referred to as 
the "normal" acoustic impedance, defined as the ratio of pressure to the compo- 
nent of particle velocity which is at right angles to the surface. The value as thus 
defined is usually independent of the angle of incidence of the sound wave, and 
when this is true, the normal impedance may be thought of as a constant or fixed 
property characterizing the acoustical structure at any given frequency. In cer- 
tain cases the normal impedance does vary with the angle of incidence, as when 
the spacing of the partitions in an air space behind a porous material is greater 
than one-half wavelength, thus allowing wave motion in the air space parallel to 
the surface. 



Vol 49, No. 3 

from point to point in magnitude and phase from the characteristic 
impedance of air. Acoustic impedance is, therefore, usually expressed 
in relation to the characteristic impedance of air, as Z/pc. 

The relation between the value of impedance existing at the surface 
of a material and the absorptivity of the surface may be derived by 
analyzing the pressure and particle velocity relations in the standing- 
wave system when the sound wave incident on the material combines 





6 8 


FIG. 1. Relation of normal-incidence absorption 
coefficient a n to acoustic impedance Z = R -\- jX. 

with the reflected wave of reduced amplitude. For the case of nor- 
mal or 90-degree incidence of sound, waves, the analysis results in the 
curves of Fig. 1, in which the absorption coefficient a n is plotted as a 
family of contours in terms of the two impedance components R/pc 
and X/pc. The following important points are evident : 

(1) A surface having a given absorption coefficient may have any 
one of an infinite number of impedance values which fall on the 



contour corresponding to that coefficient. In other words, the 
absorption coefficient does not uniquely determine the impedance, 
but a given impedance does uniquely determine the absorption. 

(2) Sound absorptivity depends on how closely the impedance of 
the absorbing surface matches the characteristic impedance of air. 
It will be noted that 100 per cent absorption is obtained only when 





- 8 








FIG. 2. Relation of random-incidence absorption 
coefficient a r to acoustic impedance, based on assump- 
tion of completely diffuse sound field. 

the impedance match is perfect; that is, when the reactance X is 
zero, and the resistance R is equal to pc. It will be shown later that 
a material surface having this impedance is physically realizable at 
single frequencies. The effect of acoustic impedance matching is 
analogous to the well-known electrical case where maximum power 
transfer results from matching the load impedance to that of the 

266 SABINE Vol 49, No. 3 

(3) Reducing the magnitude of either positive or negative re- 
actance always increases absorptivity and produces a maximum of 
absorption at zero reactance. The absorption is always zero, 
however, at zero acoustic resistance. The physical interpretation of 
this is the important fact that sound energy can only be absorbed, 
in the true sense of being transformed into heat energy ,^by encoun- 
tering a frictional resistance in the material. 

When sound-absorbing material is placed in a room, the sound 
waves strike it not only at right angles, but randomly at all angles. 
The absorptivity of a material averaged for all angles of incidence is 
substantially higher than for normal incidence. This is shown* by 
Fig. 2 where contours similar to those in Fig. 1 are drawn for the case- 
where the sound field is completely diffuse and sound waves are 
striking the material at all possible angles simultaneously. This con- 
dition is essentially attained in rooms having dimensions large com- 
pared to the wavelength of sound and containing average surface ir- 
regularities such as normal furnishings and architectural details. In 
the reverberation chambers in which the published absorption co- 
efficients of commercial materials are measured, the diffuse con- 
ditions assumed in Fig. 2 are approximated at high frequencies, but 
at the middle and low frequencies, limited data indicate that the meas- 
ured values are higher than wbuld be predicted from the absorption- 
impedance relation of Fig. 2. Further investigation on this subject 
is necessary. It will be noted further from Fig. 2. that 100 per cent 
absorptivity at random incidence is obtained with a value of R/pc 
slightly higher than unity, namely, about 1.4. 


Having examined the relation of impedance to absorption, we are 
now interested in determining the connection between the impedance 
and the physical properties of an acoustical material and its mounting. 
Exact formulas have been worked out mathematically 1 ' 4 for certain 
cases which give impedance in terms of physical constants, of which 
some are accurately measurable and others must be assigned esti- 
mated values. Experimental work has shown that fairly accurate 
checks with the theoretical predictions can be made from knowledge 
of only those physical properties which are directly measurable. In 
the case of a homogeneous, porous material mounted directly against 
a rigid backing, these properties are (7) thickness; (2) porosity P, 

* See Fig. 30, p. 141, of Reference 1. 



defined as the ratio of the volume of air in the pores of the material 
to the total volume; and (3) specific flow resistance r, defined as the 
frictional resistance to direct-current air flow offered by unit thick- 
ness and area of the material. The flow resistance is determined by 
the size and configuration of the pores and the fibers or particles form- 
ing them. 


The case just referred to is the simplest type of sound-absorbing 
structure, and being also the most instructive, will be analyzed in 
some detail. Fig. 3 shows the measured acoustic impedance values 


512 1024- 



FIG. 3. Measured impedance of one-inch homogeneous porous materials 
having indicated values of specific flow resistance r. Samples tested against 
rigid backing. 

of a series of porous materials all one inch thick, and having varying 
values of specific flow resistance. The two impedance components 
are plotted separately against frequency, each pair of R and X curves 
representing a single material having the value of flow resistance indi- 
cated. The samples are all fibrous, but of varying density and basic 
material, including glass wool, rock wool, and vegetable fiber. The 
following points may be observed from these curves : 



Vol 49, No. 3 

( 1 ) The acoustic resistance R/pc is nearly independent of frequency 
and is roughly proportional to the direct-current flow resistance of the 

(2) The acoustic reactance X/pc is negative, approximately in 
inverse proportion to the frequency, and nearly independent of the 
flow resistance of the material. 

(3) A similar set of curves plotted for a different thickness would 
show that the acoustic 'resistance increases with the thickness, and 
that the reactance varies inversely with the thickness. 

The physical reasons for these impedance characteristics may be 
better understood by setting up a mechanical model or analog repre- 
senting the action of sound pressure on a porous material, as in Fig. 4. 


R = l /*rL. X = - 




FIG. 4. Mechanical and electrical analogs of sound 
absorption by a homogeneous porous material of thick- 
ness L, porosity P, and specific flow resistance r. 

If the distance L from the surface of the material to its rigid backing, 
namely, the thickness, is smaller than one-quarter wavelength, the 
motions of all air particles within the material are essentially in phase 
with each other, and the body of air acts as a simple cushion or spring. 
The alternating sound pressure at the surface may be represented by 
a vibrating weightless piston. The spring, as is well known, presents 
a'stiffness reactance against the motion of the piston which is nega- 
tive (velocity leads force), and inversely proportional to the fre- 
quency. Reducing the thickness L would be equivalent to making 
the spring shallower and therefore stiffer, thus increasing the re- 
actance. In addition to the stiffness of the air cushion, the air mov- 
ing in the pores of the material encounters frictional resistance, or 
damping, which is essentially independent of frequency and whose 
value is determined by the pore structure of the particular material. 
The resistance is represented in the mechanical model by the points 


attached to the spring coils engaging the serrated block. The re- 
sistance varies directly with the thickness, or, in the mechanical model, 
with the number of spring coils contacting the resistance block. The 
frictional resistance and the stiffness reactance in series make up the 
acoustic impedance of the porous material, or the mechanical im- 
pedance of the model. The electrical analog of a capacitor and re- 
sistor in series is also shown. 

At low frequencies, the impedance of the simple structure shown in 
Figs. 3 and 4 may be expressed by the following approximations: 

R ^ 1 / z rL 
YC*- PC * 


Experimental deviations from these values have been observed and 
attributed to differences between the static and dynamic values of 
r and P. In Fig. 3, the fact that with increasing frequencies the 
resistance R/pc deviates from a constant value and that the react- 
ance X/pc changes with the resistance is due to the effects of the 
wavelength becoming smaller in relation to the thickness. 

We can now observe how the physical properties of materials are 
related to sound absorption through impedance by replotting the im- 
pedance-frequency curves of Fig. 3 and superimposing them on the 
impedance-absorptivity contours of Fig. 2. This is done in Fig. 5. 
Each point represents the measured acoustic resistance and reactance 
values of a single material at one of the four frequencies indicated, the 
random-incidence absorption coefficient a T for each point being given 
by its position with respect to the absorption contours. The points 
joined by each vertical line indicate the frequency characteristic of 
the particular material having the indicated direct-current flow re- 
sistance r, and the horizontal lines denote the variation of absorption 
with flow resistance of the various materials at each frequency. As 
before, all materials are one inch thick. 

From these curves we may derive several useful facts bearing on 
the practical design of acoustical materials. The securing of high 
absorption depends on the right combination of both resistance and 
reactance. In other words, either component alone may be re- 
sponsible for limiting the absorptivity. For example, knowing that 
the reactance X/pc is governed principally by thickness and fre- 
quency, we see from the curves that a one-inch material cannot be 
expected to have an absorption coefficient at 256 cycles of greater than 
about 0.35 regardless of its porous structure or flow resistance. 



Vol 49, No. 3 

Doubling the frequency to 512 cycles reduces the reactance by ap- 
proximately one half, and the maximum absorption attainable at this 
frequency is raised to about 0.53. The same effect could be 'obtained 
by doubling the thickness instead of the frequency. 

It will be noted further that because of the positions of the absorp- 
tion contours, the absorption peak occurs at progressively lower values 
of flow resistance with increasing frequencies. In other words, a ma- 
terial of a given thickness cannot have a value of flow resistance such 











FIG. 5. Data of Fig. 3 shown in relation to contours 
of random-incidence absorption coefficient <x r . 

that the maximum absorption attainable for that thickness is reached 
at both the high and low frequencies at the same time. This can be 
shown better by plotting the experimental data of Fig. 5 (plus data 
for a few additional samples) as curves of absorption coefficient vs. 
flow resistance for each frequency. This is done in Fig. 6, where each 
vertical row of points represents a separate material. The average 
coefficient for the particular four frequencies chosen is defined as the 
Noise-Reduction Coefficient, and its curve is also shown. Materials 


with very low flow resistance, such as low-density glass or rock wool, 
show the widest variation between low- and high-frequency absorp- 
tion, with less than maximum average absorption. As the flow re- 
sistance is increased, the spread between low and high frequencies is 
reduced and the average absorption rises to a broad peak. With 
further increases of flow resistance, as in high-density, relatively non- 
porous boards or tiles, the absorption-frequency characteristic be- 
comes progressively flatter, but at the expense of over-all absorptivity. 
The effect of porosity P, defined previously as the ratio of air vol- 
ume to total volume of a porous material, was not considered in the 




120 160 2OO 240 




FIG. 6. Relation of random-incidence absorption coefficient a r to specific 
flow resistance of one-inch homogeneous porous materials on rigid backing. 
Data taken from impedance measurements. 

foregoing analysis of experimental data. Since the stiffness of the 
air in the pores of a material depends on the air volume alone rather 
than the gross volume of air and material, the reactance is determined 
by the "effective" thickness which is the actual thickness multiplied 
by the porosity. Measurements by Beranek 4 have shown that prac- 
tically all fibrous materials having usable sound absorption have 
porosities of more than 85 per cent and that the porosity varies much 
less between materials than the flow resistance. In the materials, all 
of them fibrous, for which the above test data were determined, the 
porosity enters only as a minor factor which does not greatly affect 
the absorption nor cause appreciable deviation of the experimental 

272 SABINE Vol'49, No. 3 

points from a smooth curve. Materials composed of very coarse 
shreds or of solid particles bonded together may have porosities rang- 
ing as low as 30 or 40 per cent, and the maximum absorption of which 
they are capable is limited by effective thicknesses which are less than 
their actual thickness by these ratios. 

Summing up the characteristics of the simplest type of sound- 
absorbing structure as heretofore analyzed, it may be stated as a first 
approximation that at low frequencies the absorption is limited by 
the reactive component of impedance which in turn is governed by 
the thickness, and that at high frequencies the absorption is controlled 
by the resistive component of impedance which depends on the flow 
resistance of the material. This explains what has been known ex- 
perimentally for many years, that increasing the thickness of a porous 
material increases the low-frequency absorption and, so to speak, pro- 
gressively lowers the low-frequency cutoff, without appreciably chang- 
ing the high-frequency absorption. 


The next simplest type of acoustical structure, and one of the most 
common, is the homogeneous porous material mounted with an air 
space between it and a rigid backing. The well-known effect of this 
mounting in increasing the low-frequency absorption may be caused 
by either or both of two factors, namely, increase of the effective thick- 
ness and diaphragmatic vibration. Neglecting the latter and assum- 
ing that there is no rigid, impervious layer which would restrict the 
vibratory motion of air through the back of the material, the inter- 
posing of the air space provides a deeper and therefore softer air 
cushion between the outer face of the material and the rigid backing. 
This results in lowered acoustic reactance and higher absorption in 
the frequency region where the absorption is reactance controlled. 
The deeper the air space, the lower the frequencies at which effective 
absorption increases occur. Thus the very practical point is brought 
out that increased absorption at low frequencies can be obtained only 
by increasing the total space between the face of the treatment and the 
rigid -backing, but that most of the space need be occupied only by 
inexpensive air rather than costly acoustical material. This is taken 
advantage of in the hanging of drapes for acoustical purposes. When 
hung in close contact with the wall they have high absorption only 
at the very high frequencies. By spacing them out a foot or two, the 
curve can be flattened out over a large part of the frequency range. 


When materials of the board-type are light enough and flexible 
enough to vibrate diaphragmatically when mounted over an air space, 
the absorption depends both on the motion of air into the pores and 
the motion of the material surface itself. The introduction of the 
three additional constants governing the diaphragmatic vibration of 
the material, namely, its mass, its flexural stiffness, including that of 
its attachment to the supporting members, and the frictional resist- 
ance to bending set up in the material and its supports, result in a 
rather complicated over-all impedance function. The limited studies 
carried out so far indicate that appreciable increases in absorption by 
diaphragmatic vibration of highly porous materials occur only at 
mechanical resonance frequencies, and then only when the mechanical 
constants are such as to produce a sharp, well-defined resonance peak. 


Diaphragmatic absorption is much simplified if the material is non- 
porous. In this case the motion of its surface, when it vibrates as a 
whole, is that of a mass supported by a spring whose total stiffness is 
the sum of that of the air cushion under the diaphragm, and the flex- 
ural stiffness of the diaphragm and its supports. The diaphragm ex- 
hibits the usual resonant frequency determined by the ratio of stiff- 
ness to mass. The acoustic impedance of a diaphragm is simply the 
mechanical impedance per unit area. At the resonant frequency the 
negative acoustic reactance caused by stiffness equals the positive 
reactance due to the mass of the diaphragm, resulting in a net acoustic 
reactance X/pc equal to zero. From the absorption-impedance re- 
lations shown in Fig. 2, we know that the absorption will reach a peak 
at the resonant frequency, and that the height of this peak will de- 
pend only on the acoustic resistance R/pc. The acoustic resistance 
in turn is given directly by the mechanical frictional resistance set 
up in the bending of the material. As in the case of porous ma- 
terials, the acoustic resistance of a diaphragm can be too low as well 
as too high for maximum possible absorption. Many attempts at 
developing vibratile acoustical materials have been made on the as- 
sumption that vibration alone is sufficient for high absorption. Ma- 
terials such as thin metal or hard paper, for example, while they may 
vibrate quite freely in a sound field, have internal bending resistances 
which are much too low to provide the proper match with the char- 
acteristic impedance of air pc. In other words, since there is not 
enough frictional resistance to transform more than a small percentage 

274 SABINE Vol 49, No. 3 

of incident sound energy into heat, the rest must necessarily remain as 
sound in a reflected wave. 

The curved panels of thin plywood, used as diffusing elements in 
radio studios and sound stages, furnish a good example of diaphrag- 
matic absorption. By varying the depth of the air spaces and the 
spacing of the supports, the resonant frequencies are staggered over a 
wide range. Still higher resonant frequencies are obtained through 
segmental vibration of the panels. The resultant average absorption 
curve is therefore quite smooth but of comparatively low value. 


One of the most common variations of the above-described ele- 
mentary types of acoustical material is the provision of a perforated 
rigid plate such as asbestos cement board over a porous material. 
From standard acoustical theory we find that for wavelengths larger 
than the dimensions or spacing of the perforations, the acoustic im- 
pedance of such a plate is essentially a pure positive reactance (with 
a negligibly small resistive component) which increases with the fre- 
quency and with the thickness of the plate and is inversely propor- 
tional to the per cent of plate area occupied by the perforations. At 
low frequencies its value is given by 

where / is the thickness of the plate, D is the diameter of the perfo- 
rations, and k is the per cent of open area. 5 

Fig. 7 shows the measured impedance and the corresponding theo- 
retical random-incidence absorption values of a one-inch sample of 
porous material having a specific flow resistance of 19, with and with- 
out a rigid facing which is 3 /ie inch thick, perforated with 3 /ie- 
inch diameter holes, 17 / 3 2 inch on centers, and spaced out 
Vie inch from the porous material. At low frequencies the positive 
reactance of the plate is so small compared to the negative reactance 
of the air in the material that the presence of the plate has no appre- 
ciable effect on the absorption. With increasing frequencies the 
positive reactance of the plate becomes increasingly large with rela- 
tion to the negative reactance of the air behind it until a resonance 
frequency is reached at which the net reactance is zero. At this fre- 
quency the absorption is considerably increased over that of the ma- 
terial without the perforated covering. It is determined by the re- 
sistive component alone, which, as has been seen before, depends on 



the flow resistance of the material. Above the resonant frequency 
the reactance increases, being controlled almost entirely by the posi- 
tive reactance of the perforated plate, and the absorption drops off. 
With frequencies higher than the range shown in Fig. 7 the wave- 
length would become comparable to the dimensions of the structure 
and the impedance could no longer be considered as due to lumped 
elements. The absorption would not continue to decrease but would 
tend to fluctuate about some low value. 





S- 2 








512 1024 2048 


FIG. 7. Measured impedance and absorption of one-inch porous material 
with and without perforated facing. 

It may be noted that at the resonant frequency the reactance is 
zero, thus fulfilling one of the requirements for 100 per cent sound 
absorptivity. The other requirement may be met simply by choos- 
ing a flow resistance of the material such that the acoustic resistance 
divided by pc equals about 1 .4 for random angle of incidence, or 1 for 
normal incidence. As mentioned earlier, perfect impedance matching 
and the resulting 100 per cent sound absorption can thus be physically 

276 SABINE Vol 49, No. 3 

If the flow resistance of the porous element is not too high nor the 
wavelength too short, the resonant frequency at which the absorption 
reaches a maximum can be calculated quite accurately from the 
dimensions of the perforated plate and the air space behind it. This 
frequency rises with increasing hole spacing and plate thickness and 
with decreasing hole diameter and depth of air space. The drop-off 
of absorption both above and below the resonant peak is quite pro- 
nounced for the particular structure shown in Fig. 7. A higher flow 
resistance would have resulted in increased absorption on both sides 
of the peak. Increasing the thickness, however, would raise the ab- 
sorption only at frequencies below resonance. 

The behavior of the structure may be understood better if it is con- 
sidered, by analogy, that the plugs of air in the perforations act to- 
gether as a mass which is supported by the spring or cushion of air 
behind the plate, forming a simple series-resonant system. (The 
electrical analogy would be the addition of an inductance in series 
with the resistance and capacitance shown in Fig. 4.) At frequencies 
below resonance, the air in the holes moves freely and the total air 
motion at the outer surface, and the resulting sound absorption, are 
restricted mainly by the stiffness of the air cushion. At resonance the 
air plugs find their natural period of vibration with the air spring, re- 
sulting in large amplitudes of air movement at the surface. The mo- 
tion is limited only by the frictional resistance of the porous material 
and the absorption reaches a maximum. As the frequency rises above 
resonance it becomes increasingly difficult for the sound pressure to 
overcome the effective inertia of the air plugs, and the air motion at 
the surface and the absorption correspondingly decrease. At high 
frequencies, therefore, the dimensions of the perforations have con- 
siderably more effect on absorption than characteristics of the under- 
lying structure. 

In practical constructions involving perforated covering plates, the 
size and spacing of the perforations are necessarily dictated by require- 
mentsof appearance, paintability, light reflection, and structural quali- 
ties, as well as by absorption characteristics. This compromise some- 
times results in losses of high-frequency absorption which may be too 
high for certain requirements. This condition may be corrected by the 
use of drapes or fabrics hung close to the wall so as to furnish high ab- 
sorption at only the high frequencies, or by substituting a perforated 
board which is either thinner or has a larger per cent of open area. 

Since the acoustical behavior of a perforated plate is equivalent (at 


low frequencies) to that of a mass, it would be expected that similar 
results could be obtained by placing an impervious membrane having 
a definite mass but negligible stiffness and internal bending resistance 
over a porous material. This proves to be the case, and in practice 
such membranes are used for various purposes. For example, an 
impervious paper wrapping over an absorbent element will prevent 
dusting and the depositing of dirt by direct air flow, but if it is made 
light enough its acoustic reactance will be so low over the usual fre- 
quency range that sound absorption will not be impaired. Heavier 
membranes such as building felt or paper can be used over porous ma- 
terials to secure resonances and absorption peaks at definite fre- 
quencies and to reduce high-frequency absorption when desired. In 
one specialized type of treatment commercially available for studios, 
a number of such membranes of varying weight are alternated with 
porous materials to secure overlapping resonance peaks and resulting 
high absorption over an unusually wide frequency range. 


Probably the most widely used general class of commercial acousti- 
cal materials is the integral porous tile having a painted or otherwise 
impervious surface with numerous openings in the form of perfo- 
rations, slots, or fissures which allow access of sound to the porous 
interior. Analysis of measurements on this type of material show 
that their action is essentially the same as that of the homogeneous 
porous material faced with a separate perforated plate as described 
above. The plate in this case, being simply the impervious coating, 
has negligible thickness, and the positive or mass reactance due to the 
openings through it can be predicted from the standard expressions 
for apertures in an infinitely thin plate. The negative or stiffness 
reactance is, as before, given by the total volume of air between the 
face and the rigid backing. The acoustic resistance depends not only 
on the flow resistance of the porous body of the material, but also to 
varying degrees on the size and depth of the openings. Moreover, 
in felted fibrous materials the fibers lie in planes parallel to the sur- 
face, with the result that the flow resistance, except for extremely 
loose, open structures, is much lower in the lateral direction than 
transversely. Absorption of sound takes place almost entirely 
through air moving in pores which are parallel to the surface and 
which communicate with the outside only through the walls of the 
perforations or slots. The acoustic resistance and the absorption, 


therefore, depend on what may be termed "lateral" flow resistance, 
as distinguished from "transverse". The lateral flow resistance in 
turn depends inversely on the interior surface area of the perforations. 
Thus the acoustic resistance of a perforated material may be con- 
trolled not only by choice of the basic fiber structure but by the 
number, depth, and diameter of the perforations or openings. Here 
again, the total range of possible absorption characteristics obtainable 
by these and other variables is to some extent limited in practice by 
considerations of appearance, maintenance, and structural qualities. 


1 MORSE, P. M., AND BOLT, R. H.: "Sound Waves in Rooms," Rev. Mod. 
Phys., 16 (April 1944), p. 69. 

2 BERANEK, L. L.: "Precision Measurement of Acoustical Impedance," /. 
Acous. Soc. Amer., 12 (July 1940), p. 3. 

3 BERANEK, L. L. : "Acoustical Impedance of Commercial Materials and the 
Performance of Rectangular Rooms with One Treated Surface," J. Acous. Soc. 
Amer., 12 (July 1940), p. 14. 

4 BERANEK, L. L.: "Acoustic Impedance of Porous Materials," /. Acous. Soc. 
Amer., 13 (Jan. 1942), p. 248. 

6 OLSON, H. F., AND MASSA, FRANK: Applied Acoustics, Second Edition, P. 
Blakiston's Son and Co., Inc., Philadelphia, Penna., pp. 33-34. 


DR. E. W. KELLOGG : There are two questions that I want to ask. Does it 
sometimes occur that a certain space, with a certain total available thickness, gives 
more absorption by a small space plus a spaced-out thickness of absorbent ma- 
terial than if you filled it up solid with absorbent material ? 

The other had to do with the absorption by panels. I was discussing it with one 
of our memberb and the question arose in our minds whether you can gauge what 
you might call the frictional coefficient for material by taking a sample and hitting 
it with your fingers or a stick and observing the duration of the ring it will give. 
Is it fair to form any judgment as a rather violent movement would be produced by 
a stick as compared with the very small movement that would take place in wall 

- MR. HALE J. SABINE: The answer to the first question is that the difference in 
absorption depends on the flow resistance of the particular material used. If thin 
material is used it generally has to have a higher specific flow resistance than a 
material which fills the entire cavity, if the absorption is to be the same. 

The answer to the second question tapping a vibratile membrane does excite its 
motion in much the same way that sound waves would. The ringing or the 
duration of the noise which is produced by the tapping is a fair measure of the 
damping properties. If it gives off a dull thud it has high damping properties. 
If it gives off a long, pronounced ring, the damping is probably too low for sound 
absorption. As a matter of fact, it would take a lot of experience correlated with 
measurements to be able to judge the absorption qualities very accurately by that 
means, but roughly it is a proper indication. 


The adverse effects on motion picture set lighting because of inade- 
quate wiring or unbalanced load distribution have not been previously 
emphasized in these reports. Under certain conditions of power dis- 
tribution, the incoming set voltage at a given lamp may vary from as 
low as 100 v to the output voltage of the generator, which is usually 
120 v, and in extreme cases the voltage on one leg of the 3- wire system 
actually may be higher than the generator output voltage. It is the 
purpose of this report to describe motion picture set power-distribu- 
tion methods and to illustrate the necessity of adequate cable capacity 
and balanced loads. 

In motion picture studio practice the main motor generators sup- 
plying power for set lighting are usually located in a central power- 
house and permanent cable is strung to the "bull" switches of the 
various sets. If the lighting load on a given set is sufficient to cause 
considerable line drop from the permanent installation, portable 
motor generators are sometimes placed outside the set to take care of 
the overload. In other cases, however, particularly where the set is 
located at a distance from the powerhouse and is cabled for average 
loads, a considerable voltage drop in the line from the powerhouse to 
the set is encountered when the lighting load is heavy. 

A much more serious problem in voltage drop is encountered on the 
set where the 3- wire is strung from the "bull" switches all over the area 
of the stage to spots where lamps have been placed. These installa- 
tions are of a very temporary nature and the load is constantly changing 
because of requirements of the director of cinematography as to light 
levels, balance, and changes from long shot to medium shot to close-up. 

An ideal condition exists where the load is known, remains con- 
stant, and the installation is carefully balanced and checked with 
proper instruments. In practice, this ideal is seldom attained. Usually 
the chief set electrician balances and rebalances the load from knowl- 
edge gained through experience. In general, he knows the current 
each lamp will draw from the line and he supervises the installation 
for equal load on each side of the 3-wire and decides upon the number 
of cables which must be paralleled to ensure a minimum of line loss. 
Too often the pressure of high-speed operation results in serious line 
loss and unbalanced loads. 

* Submitted Oct. 25, 1946. 




Two types of circuits are used in motion picture set lighting for 
direct-current load. 

(1) A simple 2-wire type, shown in Fig. 1, which in actual studio 
practice is found only on locations where the power is supplied from 
2-wire, gas-engine-driven generators. 

200 FEET 



life VOLTS 

FIG. 1. Single 4/0 feeders connected in series. 

(2) By far the largest part of the distribution of power is by the 
Edison 3- wire system, a diagram of which is shown in Fig. 2. In 
this system two generators are connected in series; that is, the 
positive terminal of one generator is connected to the negative 
terminal of a second generator. The point at which the generators 
are connected is called a "neutral." In this circuit each generator 
delivers its normal voltage to the load between the neutral and the 
"outside" leg as shown in the diagram. 


FIG. 2. 


Balanced circuit, single 4/0 feeders connected 
in series. 

In this discussion the current will be considered to flow in accord- 
ance with the electron theory which is that the current or electrons 
flow from the negative terminal through the load and back to the 
generator through the positive terminal. 

Where the load is exactly balanced no current flows in the neutral, 
but where the load is unbalanced the current through the neutral is 
equal to the difference between the current in the two outside legs. 



In the case of the balanced load, as outlined above, the current 
would flow from the negative terminal of generator No. 1 through the 
negative lead, through load No. 1, and then through the cable con- 
necting load No. 1 to load No. 2, through load No. 2 back to the 
positive terminal of generator No. 2. Should the circuit become un- 
balanced, as shown in Fig. 3, then all of the current would flow through 
the negative lead, through load No. 1 to the neutral point, where part 
of the current would flow back to the neutral point between the two 
generators, and then to generator No. 1. The balance of the current 
would flow through load No. 2 and back to the positive terminal of 
generator No. 2. 

It is possible in extreme conditions of unbalance to have the voltage 
applied to one load greater than the generator voltage. This can 

200 FEET 

300 AMPS 

24-O V 




inn AM P^ 1 T 



. -* 2OOAMPS [ 

NO. 2 

FIG. 3. Unbalanced circuit, single 4/0 feeders 
connected in series. 

happen where the loss in. the neutral is greater than the loss in the out- 
side conductor which is carrying the least load; a condition which is 
sometimes found on sets where a single neutral is used on a long run 
and the outside lines have been doubled up by using several extra 
conductors. Cases have been known where the voltage on one side of 
the line at the set was 15 v above that at the generator, causing incan- 
descent lamps to burn out and carbon arcs to become unstable from 

This condition may probably become clearer if we consider the cir- 
cuit shown in Fig. 3 with the neutral disconnected. We would then 
have load No. 1 and load No. 2 with the two outside leads connected 
in series across the 240 v of the two generators. 

By the use of Ohm's law we find the resistance of load No. 1 to be 
0.387 ohm and the resistance of load No. 2 is 0.595 ohm. To these 
figures should be added the resistance of the cables which would then 
be 0.01 + 0.01 + 0.387 + 0.595 = 1.002 ohms total resistance. The 



Two types of circuits are used in motion picture set lighting for 
direct-current load. 

(1) A simple 2- wire type, shown in Fig. 1, which in actual studio 
practice is found only on locations where the power is supplied from 
2- wire, gas-engine-driven generators. 

200 FEET 



FIG. 1. Single 4/0 feeders connected in series. 

(2) By far the largest part of the distribution of power is by the 
Edison 3- wire system, a diagram of which is shown in Fig. 2. In 
this system two generators are connected in series; that is, the 
positive terminal of one generator is connected to the negative 
terminal of a second generator. The point at which the generators 
are connected is called a "neutral." In this circuit each generator 
delivers its normal voltage to the load between the neutral and the 
"outside" leg as shown in the diagram. 


NO. 2 

FIG. 2. Balanced circuit, single 4/0 feeders connected 
in series. 

In this discussion the current will be considered to flow in accord- 
ance with the electron theory which is that the current or electrons 
flow from the negative terminal through the load and back to the 
generator through the positive terminal. 

Where the load is exactly balanced no current flows in the neutral, 
but where the load is unbalanced the current through the neutral is 
equal to the difference between the current in the two outside legs. 



In the case of the balanced load, as outlined above, the current 
would flow from the negative terminal of generator No. 1 through the 
negative lead, through load No. 1, and then through the cable con- 
necting load No. 1 to load No. 2, through load No. 2 back to the 
positive terminal of generator No. 2. Should the circuit become un- 
balanced, as shown in Fig. 3, then all of the current would flow through 
the negative lead, through load No. 1 to the neutral point, where part 
of the current would flow back to the neutral point between the two 
generators, and then to generator No. 1. The balance of the current 
would flow through load No. 2 and back to the positive terminal of 
generator No. 2. 

It is possible in extreme conditions of unbalance to have the voltage 
applied to one load greater than the generator voltage. This can 



FIG. 3. Unbalanced circuit, single 4/0 feeders 
connected in series. 

happen where the loss in. the neutral is greater than the loss in the out- 
side conductor which is carrying the least load; a condition which is 
sometimes found on sets where a single neutral is used on a long run 
and the outside lines have been doubled up by using several extra 
conductors. Cases have been known where the voltage on one side of 
the line at the set was 15 v above that at the generator, causing incan- 
descent lamps to burn out and carbon arcs to become unstable from 

This condition may probably become clearer if we consider the cir- 
cuit shown in Fig. 3 with the neutral disconnected. We would then 
have load No. 1 and load No. 2 with the two outside leads connected 
in series across the 240 v of the two generators. 

By the use of Ohm's law we find the resistance of load No. 1 to be 
0.387 ohm and the resistance of load No. 2 is 0.595 ohm. To these 
figures should be added the resistance of the cables which would then 
be 0.01 + 0.01 + 0.387 + 0.595 = 1.002 ohms total resistance. The 



total current which would flow through the circuit under this condi- 
tion would be 240/1.002 = 239.5 amp. Therefore the voltage drop 
across load No. 1 would be 239.5 X 0.387 = 92.7 v. The voltage 
drop across load No. 2 would be 239.5 X 0.595 = 1425 v. 


FIG. 4. Two-wire generator single 4/0 feeders. 
Voltage at the set. 



Distance in Feet 



















FIG. 5. Two-wire generator double 4/0 feeders. 
Voltage at the set. 



Distance in Feet 


















-The foregoing would be the result if the neutral were disconnected 
as mentioned above. If the neutral were reconnected to the circuit 
through a high variable resistance there would be little change in the 
condition shown. However, if the resistance in the neutral were 
lowered until reduced to the resistance of 200 ft of 4/0 cable, the cir- 
cuit would be brought back to the conditions shown in Fig. 3. 

The line loss in a 3-wire circuit is much less than in a 2-wire circuit 
carrying the same load, which will be seen from the following example. 
If we were to connect the 600-amp load shown in Fig. 2 to a 2-wire 



circuit of single 4/0 cable (this would be much beyond the capacity of 
single 4/0 and is used merely for illustration), then the voltage at the 

120 vi 

FIG. 6. Two- wire generator triple 4/0 feeders. 
Voltage at the set. 


Distance in Feet 

































































































120 VI 

FIG. 7. 

Three-wire generators single 4/0 feeders. 
Voltage at the set. 

Amperes at 240 v 


Distance in Feet 


















load would be as follows : The total resistance of the line would be 0.02 
ohm and the total current 600 amp. The line loss would then be 600 
X 0.02 = 12 v. Compare this to the line loss of 3 volts shown in Fig. 2. 
Another important principle is that the line loss is reduced in pro- 
portion to the square of the increase in voltage. In the 3- wire cir- 
cuits given above, while we have 120 v at the generators and apply 
this voltage to the load, the actual voltage of transmission is 240 v. 
Figs. 4 to 9 show voltages set under varying load conditions. Table 1 





120 VI 

FIG. 8. Three-wire generators double 4/0 feeders. 
Voltage at the set. 


Distance in Feet 


















shows the carrying capacity of the copper wire ordinarily used in 
motion picture studio set lighting. 



FIG. 9. Three-wire generators triple 4/0 feeders. 
Voltage at the set. 


Distance in Feet 









100 - 



























































































Light Output. Table 2 shows that an extreme case of a drop from a 
normal "on-set" voltage of 115 to 100 v would theoretically result in 


a light loss to 55 per cent of normal. While an increase of from 115 
to 125 v would theoretically result in a gain to 125 per cent of the 
light, such an extreme change would cause severe unsteadiness be- 
cause of the rise in current from 149 to 170 amp. 

However, in actual practice the variation in light is much less than 
indicated by Table 2. In motion picture studio set lighting carbon- 
arc lamps, the positive-negative carbon feed ratio is fixed and is based 
on an average of 115 line volts at the lamp ballast. The arc-control 


Copper Wire 

Ampere Rating 

Size Dia in Ohms per Manufacturers 

AWG Mils 1000 Ft Underwriters Type RH 

4/0 460 0.050 225 358 

2/0 365 0.078 150 267 

2 258 0.156 90 170 

4 204 0.248 70 125 

6 162 0.395 50 94 

['8 128 0.628 35 69 

10 102 1.000 25 50 

12 80 1.588 20 37 

14 64 2.525 15 29 


Effect of Varying the Line Voltage 

NOTE: In this test a Type 170 high-intensity arc. was operated with its trim 
at all times in normal relation, i. e., with l 6 /ie in. protrusion and */2 in. gap. The 
voltage was varied and readings were taken at the conclusion of a 3-min burn- 
ing period at each successive voltage. 

Line volts decreased 
Line volts increased 

* This is normal operation. 

motors are nstalled in such a manner that when the lamps are in 
operation, the motors are energized by current at arc voltage. If the 
lamp is undervolted the burning rate of the carbons will decrease 
and they will tend to feed together, but as they approach each other 
the arc is shortened and the arc voltage drops. Since the motor is 
energized at arc voltage, it will rotate more slowly on its reduced volt- 
age and lower the feed rate. In this manner a balance automatically 

e Volts 


Arc Volts 



























is attained between carbon burning rate, arc voltage, feed motor 
speed, and carbon feed rate, providing the voltage to the lamp is held to 
close limits. If the voltage to the lamp varies more than approxi- 
mately 5 v, a number of difficulties are encountered. 

Low line voltage will result in less current flowing through the arc 
and inasmuch as the burning rate of the positive carbon decreases 
with lower current in greater proportion than the burning rate of the 
negative carbon, the protrusion of the positive carbon will increase 
unless a manual adjustment is made. Table 3 shows the effects of 
varying protrusion. 


Effect of Varying the Protrusion 

NOTE: A M-R Type 170 high-intensity arc was used on 115 line volts. The 
negative was kept in the normal position, i. e., that which it assumed with a nor- 
mal l /z in. gap and ! 6 /i 6 in. protrusion. The positive carbon was successively 
moved and allowed to burn 3 min in each position. 

Protrusion Current Arc Volts Light 

<130 68 77 

iVie 138 65 84 

143 65 94 

148 67 100 

162 64 104 

Protrusion increased < 1 Vie 167 62 95 

177 59 77 

% ~ 

* This is normal operation. 

Under conditions of low current resulting from low line voltage, the 
negative carbon will burn with a blunt point and the arc will tend to 
wander and become unstable. 

In the case of high current caused by high line voltage, the negative 
carbon will tend to spindle and the arc will flicker from overload. 

Cases have been noted where negative carbons were burning with 
blunt points in some lamps and were spindling in other lamps on the 
same set because of conditions of extreme unbalance on the 3-wire set 

It is to be noted that the variations mentioned show up only under 
conditions of extreme change. The automatic features of the lamps 
will compensate for slight variations and the lamp operator is able to 
control the current to some extent with manual adjustment. 
Nevertheless, the variation in light output and steadiness of light 
indicate that close line- voltage control will result in vastly improved 




Table 4 shows that an extreme case of drop from a normal "on-set" 
voltage of 115 to 100 v would result in a light loss to 62 per cent of 
normal. While an increase from 115 to 120 v would cause a gain to 
116 per cent of the light, such an extreme change would result in 
obtaining only 57 per cent of expected life. The foregoing is based on 
lamps with rated voltage of 1 15 v. 


Variation of Light, Life, and Wattage of Gas-filled Incandescent Lamps When 
Operated at Voltages Above or below Their Rated Voltage 


Per Cent 

Per Cent 

Per Cent at 
Watts Socket 

Per Cent Per Cent 
Light Life 

Per Cent 

For 120-V Lamps 





















































































For 115-V Lamps 









































105 ' 













































Because of changes in color temperature with voltage changes on 
both carbon arcs and incandescent lamps, the demands for proper 
voltage regulation on color sets are greater than on black and white. 
Even so, it is possible to tolerate greater differences in a 3-strip system 
such as used by Technicolor than with a monopack system where the 



prints are also made on a monopack film. If and when this latter 
printing system comes into studio use, the demands for color temper- 
ature control will increase, and with it the further need for close set 
voltage control. 


i i g 2 * 











g.' 8 

< 95 







^ 0/1 




I is 18 S 1 




















FIG. 10. Variation of color temperature with change of 
voltage. 100- to 130-v types incandescent lamps. 

Fig. 10 shows variation in color temperature with voltage for incan- 
descent lamps. The question of color temperature change with volt- 
age for carbon arcs has been previously covered. 1 


Close voltage control on motion picture sets results in economy of 
operation, greater lamp efficiency, and less manual adjustments of 
carbon-arc units. Where the studios are equipped with adequate 
wiring and where 3-wire balance is maintained by "on-set" volt- 
meters, the entire problem of set lighting is greatly simplified. 

C. W. HANDLEY, Chairman 





1 BOWDITCH, F. T., AND DOWNES, A. C. : "Spectral Distributions and Color 
Temperatures of the Radiant Energy from Carbon Arcs Used in the Motion Pic- 
ture Industry." J- Soc. Mot. Pict. Eng., XXX, 4 (Apr. 1938), p. 400. 


(Correct to Sept. 12, 1947) 

ADMISSIONS. To pass upon all applications for membership, applications for transfer, and 
to review the Student and Associate membership list periodically for possible transfers to the 
Associate and Active grades, respectively. The duties of each committee are limited to applica- 
tions and transfers originating in the geographic area covered. 


D. B. JOY, Chairman 

30 East 42d St. 
New York 17, N. Y. 



HERBERT GRIFFIN, Chairman r3 " 

133 E. Santa Anita Ave. 
Burbank, Calif. 



BOARD OF EDITORS. To pass upon the suitability of all material submitted for publica- 
tion, or for presentation at conventions, and publish the JOURNAL 

A. C. DOWNES, Chairman 

2181 Niagara Dr. 
Lake wood 7, Ohio 




CINEMATOGRAPHY. To make recommendations and prepare specifications for the 
operation, maintenance, and servicing of motion picture cameras, accessory equipment, studio 
and outdoor set lighting arrangements, camera technique, and the varied uses of motion picture 
negative films for general photography. 

J. W. BOYLE, Chairman 

1207 N. Minsfield Ave. 
Hollywood, Calif. 


COLOR. To make recommendations and prepare specifications for the operation, mainte- 
nance, and servicing of color motion picture processes, accessory equipment, studio lighting, 
selection of studio set colors, color cameras, color motion picture films, and general color photog- 

J. A. BALL, Chairman 

12720 Highwood St. 
Los Angeles 24, Calif. 





CONVENTION. To assist the Convention Vice-President in the responsibilities pertaining 
to arrangements and details of the Society's technical conventions. 

W. C. KUNZMANN, Chairman 
Box 6087 
Cleveland 1, Ohio 



* Advisory Member. 



EXCHANGE PRACTICE. To make recommendations and prepare specifications on the 
engineering or technical methods and equipment that contribute to efficiency in handling and 
storage of motion picture prints, so far as can be obtained by proper design, construction, and 
operation of film handling equipment, air-conditioning systems, and exchange office buildings. 

(Under Organization) 

FELLOW AWARD. To consider qualifications of Active members as candidates for eleva- 
tion to Fellow, and to submit such nominations to the Board of Governors. 

D. E. HYNDMAN, Chairman 

342 Madison Ave. 
New York 17, N. Y. 





FILM PROJECTION PRACTICE. To make recommendations and prepare specifications 
for the operation, maintenance, and servicing of motion picture projection equipment, projection 
rooms, film-storage facilities, stage arrangement, screen dimensions and placement, and main- 
tenance of loudspeakers to improve the quality of reproduced sound and the quality of the 
projected picture in the theater. 

G. T. LORANCE, Chairman 
63 Bedford Road 
Pleasantville, N. Y. 












HISTORICAL AND MUSEUM. To collect facts and assemble data relating to the historical 
development of the motion picture industry, to encourage pioneers to place their work on record 
in the form of papers for publication in the JOURNAL, and to place in suitable depositories equip- 
ment pertaining to toe industry. 

(Under Organization) 

HONORARY MEMBERSHIP. To diligently search for candidates who through their 
basic inventions or outstanding accomplishments have contributed to the advancement of the 
motion picture industry and are thus worthy of becoming Honorary members of the Society. 

H. W. REMERSCHEID, Chairman 

716 N. LaBrea St. 
Holly wood, 'Calif. 



JOURNAL AWARD. To recommend to the Board of Governors the author or authors of 
the most outstanding paper originally published in the JOURNAL during the preceding calendar 
year to receive the Society's Journal Award. 

J. I. CRABTREE, Chairman 

Research Laboratories 
Eastman Kodak Company 
Rochester 4, N. Y. 



* Advisory Member. 

Sept. 1947 



LABORATORY PRACTICE. To make recommendations and prepare specifications for the 
operation, maintenance, and servicing of motion picture printers, processing machines, inspec- 
tion projectors, splicing machines, film-cleaning and treating equipment, rewinding equipment, 
any type of film-handling accessories, methods, and processes which offer increased efficiency 
and improvement in the photographic quality of the final print. 

H. E. WHITE, Temporary Chairman 
342 Madison Ave. 
New York J7, N. Y. 













MEMBERSHIP AND SUBSCRIPTION. To solicit new members, obtain nonmember sub- 
scriptions for the JOURNAL, and to arouse general interest in the activities of the Society and its 

L. E. JONES, Chairman 

427 West 42d St. 
New York 18, N. Y. 









0. F. NEU 





16 Mm 


G. C. MISENER, Chairman 

6424 Santa Monica Blvd. 
Hollywood 38, Calif. 


C. R. WOOD, SR. 






NOMINATIONS. To recommend nominations to the Board of Governors for annual election 
of officers and governors. 

E. A. WILLIFORD, Chairman 

230 Park Ave. 
New York 17, N. Y. 






Vol 49, No. 3 

STUDIO LIGHTING. To make recommendations and prepare specifications for the 
operation, maintenance, and servicing of all types of studio and outdoor auxiliary lighting 
equipment, tungsten light and carbon-arc sources, lighting-effect devices, diff users, special light 
screens, etc., to increase the general engineering knowledge of the art. 

C. W. HANDLEY, Chairman 

I960 West 84th St. 
Los Angeles 44, Calif. 





TELEVISION. To study the television art with special reference to the technical inter- 
relationships of the television and motion picture industries, and to make recommendations 
and prepare specifications for equipment, methods, and nomenclature designed to meet the 
special problems encountered at the junction of the two industries. 

D. R. WHITE, Chairman 




Red path Laboratories 

E. I. du Pont de Nemours & Co. 

Parlin, N. J. 



TELEVISION PROJECTION PRACTICE. To make recommendations and prepare speci- 
fications for the construction, installation, operation, maintenance, and servicing of equipment 
for projecting television pictures in the motion picture theater, as well as projection room 
arrangements necessary for such equipment, and such picture-dimensional and screen-charac- 
teristic matters as may be involved in high-quality theater television presentation. 

P. J. LARSEN, Chairman 

1401 Sheridan St., N. W. 
Washington 11, D. C. 

F. E. CAHILL, JR., Vice-Chairman 

321 West 44th St. 
New York 18, N. Y. 



JAMES FRANK, JR., Secretary 

356 West 44th St. 
New York 18. N. Y. 









TEST FILM QUALITY. To supervise, inspect, and approve all print quality control ol 
sound and picture test films prepared by any committee on engineering before the prints are 
released by the Society for general practical use. 

F. S. BERMAN, Chairman 
111-14 76th Ave. 
Forest Hills, L. I., N. Y. 



* Advisory Member. 

Sept. 1947 



mendations and prepare specifications on engineering methods and equipment of motion picture 
theaters in relation to their contribution to the physical comfort and safety of patrons, so far 
as can be enhanced by correct theater design, construction, and operation of equipment. 


1501 Broadway 
New York 18, N. Y. 







* Advisory Member: 







*JAMES FRANK, JR., Chairman 
*F. E. CAHILL, JR., Past-Chairman 
*H. E. WHITE, Secretary-Treasurer 



*A. SHAPIRO, Chairman 

*ROBERT E. LEWIS, Secretary-Treasurer 



*W. V. WOLFE, Chairman 
*H. W. MOYSE, Past-Chairman 
*S. P. SOLOW, Secretary-Treasurer 



* Term expires Dec. 31, 1947; ** Term expires Dec. 31, 1948. 


American Standards Association: 
Sectional Committees on: 

Standardization of Letter Symbols 
and Abbreviations for Science and 
Engineering, Z10 

Motion Pictures, Z22 

C. R. KEITH, Chairman 
A. N. GOLDSMITH, Honorary Chairmman 


Acoustical Measurements and Ter- 
minology, Z24 

Photography, Z38 


Standards Council, ASA Member- 


American Documentation Institute 

Intersociety Color Council 

R. M. EVANS, Chairman 


National Fire Protection Association 

Radio Technical Planning Board 


t Alternate. 


The editors present for 'convenient reference a list of articles dealing with subjects 
cognate to motion picture engineering published in a number of selected journals. 
Photostatic or microfilm copies of articles in magazines that are available may be 
obtained from The Library of Congress, Washington, D. C. t or from the New York 
Public Library, New York, N. Y., at prevailing rates. 

American Cinematographer The Devry "1200" Series 35-Mm 

28, 7 (July 1947) Sound-Film Projection Equipment 

Historical Development of Sound (p. 12) E. W. D'Ancv 

Films Pt. 1 (p. 235) New Acetate Color Film Releases 

E. I. SPONABLE Induce Varying Opinion Anent 

Two Worlds in Technicolor (p. 236) Quality (p. 19) 

H. I. LIGHTMAN Motiograph's Wide-Latitude Shoe 

25 Years of the "Lubitsch Touch" Tension Adjustment (p. 24) 

in Hollywood (p. 238) E. WIENKE 

Flashtubes'Tp- 240) Kinematograph Weekly (British Stu- 

D W PRIDEAUX dio Section Supplement) 

Motion Picture Institute in the 365, 2097 (July 10, 1947) 

Soviet Union (p. 242) S. Ross Practical Research by Nettleford 

A Practical Combination Sunshade- Workers (p. vii) 

Matte Box (p. 256) A. LINKO M ' G - M British Studios at Elstree 

(p. xv) 

International Projectionist New Technical Developments 

22, 7 (July 1947) (p. xix) R. H. CRICKS 

RCA's Novel Periscope Projection Colour: Future Problems (p. xxxi) 

Setup (p. 10) E. STANKO T. T. BAKER 



Because of increased printing costs, the Society of Motion Picture 
Engineers is no longer able to furnish gratis copies of papers to 
authors. Beginning with this issue, authors will be billed for all 

Each author will, however, continue to receive three compli- 
mentary copies of the issue in which his paper is published. 



CAMERAMAN: Honorably discharged U. S. Army photographer, 
desires to re-enter industrial, educational production with independent 
producer or studio. Experienced in 35- and 16-mm color and black 
and white. References and complete record of experience available. 
I am an active member of the Society of Motion Picture Engineers. 
Write, wire, or telephone Charles N. Arnold, P. O. Box 995, Peoria, 111. 
Telephone 3-9865. 


Vol 49 OCTOBER 1947 No. 4 


American Films Abroad ORTON H. HICKS 297 

The Processing of Two-Color Prints by Deep-Tank 
Methods JOHN G. STOTT 306 

Current Black-and- White Duplicating Techniques Used 
in Hollywood 


Lead-Sulfide Photoconductive Cells for Sound Repro- 
duction R. J. CASHMAN 342 

Television Studio Lighting W. C. EDDY 334 

Lead-Sulfide Photoconductive Cells f< 

Magnetic Sound for 8-Mm Projection 


Synchronized 16-Mm Sound and Picture for Projection 
at 16 Frames per Second GEORGE E. H. HANSON 357 

The Optimum Width of Illumination of the Sound 
Track in Sound-Reproducing Optics 


A Photoelectric Film Cuing System IRWIN A. MOON 364 

Space Acoustics JAMES Y. DUNBAR 372 

Motion Picture Research Council 389 

Society Announcements 390 

Copyrighted, 1947, by the Society of Motion Picture Engineers, Inc. Permission to republish 
material from the JOURNAL must be obtained in writing from the General Office of the Society. 
The Society is not responsible for statements of authors or contributors. 

Indexes to the semiannual volumes of the JOURNAL are published in the June and December 
issues. The contents are also indexed in the Industrial Arts Index available in public libraries. 











^President: LOREN L. RYDER, 

5451 Marathon St., Hollywood 38. 
** Past-President: DONALD E. HYNDMAN, 

342 Madison Ave., New York 17. 
** Executive Vice-President: EARL I. SPONABLE, 

460 West 54th St., New York 19. 
"^Engineering Vice-President: JOHN A. MAURER, 

37-01 31st St., Long Island City 1, N. Y. 
** Editorial Vice-P resident: CLYDE R. KEITH, 

233 Broadway, New York 7. 
^Financial Vice-President: M. RICHARD BOYER, 

E. I. du Pont de Nemours & Co., Parlin, N. J. 
** Convention Vice-President: WILLIAM C. KUNZMANN, 

Box 6087, Cleveland 1, Ohio. 
** Secretary: G. T. LORANCE, 

63 Bedford Rd., Pleasantville, N. Y. 
^Treasurer: E. A. BERTRAM, 

850 Tenth Ave., New York 19. 

**JOHN W. BOYLE, 1207 N. Mansfield Ave., Hollywood 38. 

*FRANK E. CARLSON, Nela Park, Cleveland 12, Ohio. 

*ALAN W. COOK, Binghamton, N. Y. 
**ROBERT M. CORBIN, 343 State St., Rochester 4, N. Y. 
**CHARLES R. DAILY, 5451 Marathon St., Hollywood 38. 
"tjAMES FRANK, JR., 356 West 44th St., New York 18. 

*JOHN G. FRAYNE, 6601 Romaine St., Hollywood 38. 
**DAVID B. JOY, 30 East 42d St., New York 17. 

*PAUL J. LARSEN, 1401 Sheridan St., Washington 11, D. C. 

*WESLEY C. MILLER, MGM, Culver City, Calif. 
**HOLLIS W. MOYSE, 6656 Santa Monica Blvd., Hollywood. 
*JA. SHAPIRO, 2835 N. Western Ave., Chicago 18, 111. 
* WALLACE V. WOLFE, 1016 N. Sycamore St., Hollywood. 

*Term expires December 31, 1947. tChairman, Atlantic Coast Section. 
Term expires December 31, 1948. TChairman, Midwest Section. 
Chairman, Pacific Coast Section. 

Subscription to nonmembers, $10.00 per annum; to members, $6.25 per annum, included in 

their annual membership dues; single copies, $1.25. Order from the Society at address above. 

A discount of ten per cent is allowed to accredited agencies on orders for subscriptions 

and single copies. 
Published monthly at Easton, Pa., by the Society of Motion Picture Engineers, Inc. 

Publication Office, 20th & Northampton Sts., Easton, Pa. 

General and Editorial Office, 342 Madison Ave., New York 17, N. Y. 

Entered as second-class matter January 15, 1930, at the Post Office at Easton, Pa. 

under the Act of March 3, 1879. 


Vol 49 OCTOBER 1947 No. 4 


Summary. This paper describes the importance of the distribution of American 
films overseas to the motion picture industry, to the national economy, and to the 
cause of world peace. 

Overseas markets for American films are of vital importance to 
you as motion picture engineers and executives. Naturally you have 
been interested in these overseas markets and overseas films for many 
years, and you were especially interested in them during the war when 
you played such a large role in bringing high standard equipment and 
films to the Armed Services. 

It was my job to distribute those films, especially the 16-mm enter- 
tainment gift prints. These were donated by the industry to the 
Army and Navy. There have been some very interesting after- 
effects resulting from those gift prints. 

By acquiring the motion picture habit, the returned veterans have 
learned to demand better pictures and more pictures. Their critical 
sense has been increased. For instance, we are now able to book to 
the small towns throughout the country good pictures on the week 
ends. Previously these towns played the horse-opera type of movie. 

So far as more pictures are concerned, there was a little item in the 
New York Times that read as follows : 

"Lewes, Delaware Tomorrow night will mark the end of another 
'blue law' of this 315-year-old town. Sunday movies will be shown for 
the first time, and if the shades of the Holland Dutch founding fathers 
who landed here in 1631 hover in disapproval over the town's one 
movie theater, no one will be surprised. Their observance of the 
Sabbath was strict enough to have lasted more than three centuries. 

* Presented Apr. 21, 1947, at the SMPE Convention in Chicago. 
'* Loew's International Corporation, New York, N. Y. 


298 HlCKS Vol 49. No. 4 

"Returning service men expressed a desire for a referendum early 
this fall. When the vote was conducted a fortnight ago an over- 
whelming majority approved." 

There has been another advantage that we found overseas. Those 
gift prints, although intended only for the Army and Navy, were 
shown to the local inhabitants wherever we happened to have troops 
with the result that the motion picture habit has been inculcated in 
them. Film companies find that the 16-mm program is progressing 
rapidly in the Far East, the Middle East, West Africa, wherever we 
had troops; whereas in South America, where we did not have troops, 
the demand is not so great. 

On the debit side we find that many of these prints have been 
stolen, or let us say "borrowed". There is great difficulty in im- 
pressing upon the layman the sanctity of a copyright. That was 
brought home to us very vividly not long ago when we received a 
letter from an Army major stationed at a local air base. He told 
about the great pleasure that his family and friends had had as a re- 
sult of bringing back from Iceland a print of "For Me and My Girl" 
with Judy Garland. He had now been ordered to Alaska and he 
did not want to take a print to Alaska that was not in the best con- 
dition. He asked if we would replace it with a new print if he sent 
it back. He got the answer and we got the print. 

Another repercussion has been in the very human failing of not 
wanting to pay money for things that have previously been given 
free. An incident in that connection happened out in the Philippines 
at a little town called Los Banos, about a hundred miles from Manila. 
The first night that we tried to show pictures there a number of 
armed bandits shot up the lobby of the theater. The exhibitor re- 
fused to run the show but was forced at the point of a gun to continue. 
After the show the bandits retreated to the hills they did not rob 
anyone. They resented the fact that anybody should be charged for 
motion pictures after three years of free shows that the Army had 
been giving them! 

My present assignment is to apply to the distribution of enter- 
tainment films the same methods we learned in the Army and Navy. 
Commercially, like all corporations, we answer to a board of direc- 
tors, but culturally we answer to the world. Mr. Arthur Loew, in 
our Basic Manual, has stated his beliefs as follows : 

"What makes this world- wide distribution so significant is the 
fact that films are the nearest thing yet perfected to a universal 


language, with undeniable cultural, political, and social force. Films 
are able to convey ideas to more people more intelligibly, accurately, 
and graphically than any other medium of communication. 

"To illustrate this point, it can be said that most persons, thanks 
to motion pictures, have mental pictures of far-off places and things 
they actually have never seen. Industrial workers who have never 
left the city know much about the farmer, and vice versa, because of 
entertainment films. 'Mrs. Miniver' taught the world more about 
the spirit of England and the temper of its people than all the text- 
books put together. 

"In this peace-seeking- aftermath of war, the most distressingly 
elusive commodity is simple understanding among nations, and it is 
precisely this commodity that motion pictures are able, in and of 
themselves, to impart. 

"This is why it is our responsibility and that of the industry to 
make motion pictures available everywhere. And this is why free- 
dom of the screen must be our guiding principle." 

We have found two barriers in reaching that goal. The first 
barrier is economic and the second barrier is cultural. We are re- 
moving the economic barrier by introducing 16-mm operations into 
these remote corners of the world where 35-mm could not operate 
profitably. This is not only because of the mobility and economy of 
16-mm equipment, but also because of the cost of transporting the 
films. In some instances overseas more has been paid for 35-mm 
film transportation than was received on the royalty for the picture 
itself. We are removing the cultural barriers in two ways. The 
farther away from the centers of civilization the greater the illiteracy, 
and, therefore, superimposed titles at the bottom of the screen are 
not of any real value because the people cannot read. We have 
overcome that by dubbing the native language onto the film itself. 

The other method is a new one and is less expensive. It is called 
narration. The sound track is tuned down at certain intervals and a 
voice in the native language comes on and explains the action and 
gives other information that will make the picture more enjoyable. 
We are doing narration in five languages, Portuguese, Arabic, Hin- 
dustani, Chinese, and Siamese. Complete lip synchronization is 
done in Spanish, French, Italian, and, before the war, German. As a 
consequence many more hundreds of thousands of people are going 
to become motion picture devotees and will be drawn into the world 
circle of neighbors. 

300 HlCKS Vol 49, No. 4 

You cannot put on a program like this without trained personnel. 
At the end of the war each of our overseas offices selected some young 
national of the country involved and he was sent to the United States 
for training. There were ordinary routine matters such as spending 
days in our exchanges and nights in our theaters, a couple of weeks 
at the studios, and a few weeks at projector factories. In addition, 
we tried to give them some extracurricular activities. For instance, 
we sent them to Washington for a week to their own embassies and to 
meet the people in the United States Government who are interested 
in motion pictures. We sent them to the University of Chicago to 
see the Encyclopedia Britannica films in order to acquire some feel- 
ing toward educational films. We usually had them travel by day 
train because we wanted them to see the country. Then they came 
to New York for the final summing up. We had an editorial writer 
from the New York Times meet with them evenings and answer their 
questions about the United States. 

These men have gone back to their own countries. What are we 
doing now to give them ammunition? For one thing we are sending 
the best possible pictures. We think Hollywood makes good pic- 
tures, and we try to send the best ones, eliminating those that may 
give an unfavorable reaction toward the United States. Of course, 
the best way to remove the possibility of such a reaction is to nip it 
at the very source by not making pictures that will give an unfavor- 
able reaction. For that purpose we have created an International 
Information Center, and it is the duty of that organization to revise 
scripts, to make suggestions that will make pictures more presentable 
overseas, and to eliminate these disquieting influences that we have 
had occasionally in the past. 

In selecting pictures there is a fine line to be drawn that requires 
nicety of judgment. Sometimes we lean over backwards, I am afraid. 
You may have seen this item in one of the papers about "The 
Grapes of Wrath". "Oslo Municipal authorities today forbade a 
showing here of the motion picture, 'The Grapes of Wrath,' because 
American distributors insisted that audiences be told the conditions 
depicted are not normal in the United States." Subsequently we 
withdrew our insistence and the film was shown without any foreword. 

There are other pictures such as "Mr. Smith Goes to Washington", 
a picture that we could send to Great Britain because there the demo- 
cratic process is understood. However it is a film that we could not 
and would not send to many other countries. 


One final thing that we do in the way of providing ammunition is 
through an arrangement with the State Department. Sixteen- 
millimeter prints are prepared in the English language and sent to 
our embassies through the courier service of the State Department. 
That enables our ambassador to invite the leading politicians and 
officials of each country to see our latest pictures. So far that serv- 
ice is operating on a weekly basis only in Russia. 

In a letter from Ambassador Smith, he describes the great value 
that has resulted from showing these pictures to a small select group 
of Russian officials and he believes the results to be of great benefit to 
the American people and to American industry. 

I can tell about Russia in ten seconds because I know only three 
things: (1) They have 17,000 theaters, practically the same number 
that we have. (2) Deanna Durbin is the most popular American 
actress. (3) Within recent years we have sold only three American 
pictures to Russia and for those three we have received a total of 
$25,000 which happens to be the same amount that those pictures 
earned in the city of Prague alone. You may well say that we should 
give our pictures to Russia. Maybe the industry would if we could 
be sure that they would be shown. However, let me reassure you that 
in no case has Russia been denied any picture because of price. We 
have always been able to agree on a price that Russia is willing to pay. 

Any pictures sold to Russia in the future will be sold through the 
Motion Picture Export Association. That is a corporation formed 
under the Webb-Pomerene Act to distribute American pictures in 
those countries where a monopoly exists and where it is impossible to 
engage in free trade. There are thirteen countries thus served: 
Holland, Germany, Poland, Austria, Hungary, Czechoslovakia, 
Rumania, Yugoslavia, Bulgaria, Russia, Japan, Korea, and the 
Netherland East Indies. 

The Motion Picture Export Association is owned by the eight 
major companies. No regard is given to the selection from any one 
individual conipany. In other words, the selection is made from 
those pictures which we think will do America the most good. Natu- 
rally we try to obtain a balance in that a certain number of dramas, 
musicals, comedies, and other pictures and we are careful to try 
to introduce new American stars because they have a great box-office 
value overseas just as in this country. 

Newsreels are distributed in native languages in every country 
that will permit us to distribute newsreels. 

302 HICKS Vol 49, No. 4 

The reaction to our pictures is what you might expect. The people 
are tired of war pictures, they want escape. In a theater in Budapest 
the picture "Casablanca" was being shown. You may recall that 
in that film the Germans start singing "Watch on the Rhine" and 
they are drowned out by the French populace singing "La Mar- 
seillaise". The audience in Budapest became so imbued with this 
spirit that they joined the French chorus and kept on singing through 
the next two reels. Finally they forced the operator to rethread and 
run the reel over again. 

The theaters overseas need new equipment, new seating arrange- 
ments, and better sanitary facilities. We need new theaters every- 
where. The standards of exhibition in many countries overseas are 
not what we have here. One of the worst examples we came across 
was in a town in Gambia. The theater has one 16-mm projector, and 
the operator is afraid of overheating. Accordingly, he charges full 
admission for running one 1600-foot reel, then stops the picture and 
says, "If you want to see the rest you can come back tomorrow night 
at the same admission." 

In a little town about 60 miles from Bogota, Colombia, we had an 
exhibitor who was paying a flat rental and we wanted to check on 
him. We drove out there one night about six o'clock. We couldn't 
see any evidence of a theater. The town was about a block long and 
had a population of about 500 people. We drove up and down and 
we saw some people going into a private home. We followed them 
and inside of the home there was a man collecting money. He had 
torn down the back of his home and built a barn and that was the 
theater. There were only about 30 people, and he was charging 
them the equivalent of $1.16 apiece. That is a higher price than we 
get in our first-run theater in Bogota. After the show we asked him 
why he was charging such a high admission and why he wasn't ad- 
vertising. He said, "I want to keep outithe riffraff." Inasmuch as 
he was paying a flat rental we couldn't argue too much, but it is our 
goal to try to bring to these remote 16-mm theaters overseas the same 
standard of exhibition that we have in this country. 

As for the number of theaters overseas according to the Depart- 
ment of Commerce's latest count there are 83,668 theaters and only 
about 20 per cent of those theaters are in the United States. In 
other words, four out of five theaters are overseas. That gives you 
some idea of the importance of our overseas market. It is important 
for three reasons. It is important to the motion picture industry, 


it is important to the national economy, and it is important to the 
national security. 

It is important to the motion picture industry because if we lose 
that 40 per cent of our income or any part of it, we are going to have 
to raise admission prices in this country, or lower the quality of our 
pictures, or decrease the number of personnel working in the industry 
and decrease the amount of money that is paid for research and other 

The overseas market is important to the national economy. As 
you know, motion pictures whether American, British, French, or 
Spanish have done more than any other medium to raise the stand- 
ard of living of civilized peoples throughout the world. More wash- 
ing machines, more radios, more automobiles, have been sold in- 
directly by motion pictures than by any other form of sales enter- 
prise or promotion. 

The overseas market is important to the national security because 
motion pictures are helping to win the peace. On that subject let 
me read what Eric Johnston recently said: "The free exchange of 
ideas is even more important than the free exchange of goods. There 
cannot be one world as long as there are any 'foreigners' in it. The 
meaning of the word 'foreign' must disappear, and with it the plural- 
ity of discordant foreign policies by which the nations are divided. 
The people of the world will cease to seem strange to one another 
only when they know each other as neighbors do. To bring them to 
such knowledge of one another is a mission which the motion picture 
is peculiarly fitted to perform." 

It would be oversimplification, of course, to maintain that free- 
dom of the screen alone can bring enduring peace. But it would be 
equally naive to believe that enduring peace can be obtained without 
freedom of the screen. A free screen can bring about world under- 
standing and world education. 

It was H. G. Wells who said, "Civilization is a race between educa- 
tion and disaster." Motion pictures will play a vital role in winning 
that race! 

->i*f ' & 


MR. 1ST D. GOLDEN: In the short space of time that Loew's International has 
been 'operating the 16-mm program abroad, approximately how many theaters 
are there in the foreign market? 

MR. ORTON H. HICKS: I wish I could answer that, but I do not have the 
figures here. We break it down according to countries. In some countries where 

304 KICKS Vol 49. No. 4 

there was already a large 16-mm operation the business has been phenomenal. 
In France there are something like four thousand 35-mm theaters and about nine 
thousand 16-mm locations. Most of those are served by mobile units. I think 
there are something like 1900 mobile units covering France. In some of our 
branches in France we are making more 16-mm shipments than we are 35-mm 
shipments. In England there are several hundred. I don't knowtheexactnumber. 
If we had to figure out the whole world, I would say that France at this stage of 
the game is probably as big as the rest of the world combined, because elsewhere 
it is in the pioneering stage and it is not in France. 

MR. GOLDEN : Has not the retarding factor been your inability to secure suffi- 
cient 16-mm equipment to carry out your program? 

MR. HICKS: That has held it up somewhat, but I should be making an alibi if 
I tried to pretend that we could go ahead much more rapidly merely because of 
equipment. The projector manufacturers have been wonderfully co-operative 
and very farsighted in realizing that this foreign market is important. They have 
made much f airer allocations to the foreign market than many other industries 
have made. 

It would be helpful in some of these countries if we did not have the dollar 
problem. In Chile it is pretty difficult to persuade anyone to import equipment 
because you have to have a license and you have to pay in American dollars, 
and how are you going to get those American dollars out of the country? That is 
holding us up in many places. As a rule, I think that we are over the equip- 
ment hurdle and I know of no place where we are seriously handicapped by 
lack of equipment. I am wrong, there are some places where the countries insist 
on not importing any equipment whatsoever. They manufacture their own. 
That is true of Great Britain, Australia, and Italy. Italy and Great Britain will 
continue to be a big problem because they are shipping so much of their manu- 
factured goods overseas in an attempt to create dollar exchange. 

MR. WILLIAM KRUSE: In regard to censorship, Louis de Rochemont gave a 
slightly opposite viewpoint as to whether it is the business of the motion picture 
industry to apply censorship or flattery to the films we send abroad. Remember, 
the British let us see "Henry VIII" without polishing up his table manners. Is it 
really the business of the industry to clean them up before we send them to the 
audiences of Europe? 

MR. HICKS: I certainly am not qualified to act as spokesman for the industry 
on that subject. Just let me give you my personal opinion. I do not think we 
have any business doing that in a country such as Great Britain or Norway. An 
interesting aftermath of the little article I read you about "The Grapes of Wrath" 
was that the Norwegians refused to let the picture be seen with the title explaining 
that the conditions were not typical of the United States, that they arose because 
of certain economic conditions that prevailed for awhile, and that they were 
corrected when called to the attention of the authorities. The Norwegians said, 
"We were a free people long before you. We think we are qualified to discern the 
difference between truth and misinformation. If you want to show the picture, 
let us form our own conclusions." I think they are right. On the other hand, 
there are many countries where the showing of "Mr. Smith Goes to Washington", 
would do us harm. The democratic process has been misrepresented in some 
countries and a picture such as this would do more harm than good. 


We had an interesting thing happen in connection with "The Grapes of Wrath". 
In Yugoslavia there was a picture being shown called "American Paradise". 
We checked the records and no such picture had ever been made in the United 
States; furthermore we were not doing business with Yugoslavia at the moment. 
At last the State Department discovered what was happening. They sent one of 
their representatives and he found the theater where this particular picture was 
playing. It opened with the title, "American Paradise", and it had the regular 
credit lines and then it broke into the picture. The picture was "The Grapes of 

I think a picture shown under that title hi Yugoslavia can do us harm. 

MR. GOLDEN: May I be permitted to say a word in connection with Mr. Kruse's 
comments? I am not going to defend the industry but I do think that the Ameri- 
can motion picture industry can use good discretion in the selection of the types of 
films that we send abroad. There are many countries abroad that would like to 
select a particular type of picture that would not show the American way of life in 
the best manner. They would want it for the purpose of propaganda. Russia 
would like to select just those pictures that fit its ideology and pictures that would 
show the Russian people the worse things of American life. No one particular 
picture that is produced can show the whole drama of American life, but if you 
take a cross section of most of the pictures produced, you will have a good, clear 
understanding of the American way of life. Unfortunately, some countries have 
set up barriers against the showing of American pictures because they are fearful 
that we may turn over to their people our ideas and our ideals. They don't want 
those, they just want to pick and select certain types of pictures. 

Many reports come across my desk and they show that some foreign govern- 
ments that have an ideology contrary to ours would like to have the right of selec- 
tion . That is the reason why Mr. Hicks told you that Russia has picked only three 
pictures in recent years and purchased them from our country. They do not 
want to show our American way of life. The American picture has always been 
one of the most saleable products in foreign markets. It was also the greatest 
medium through which we sold America to the foreign countries. That is why it 
was the first commodity to be barred in foreign countries such as Italy and 
Germany long before we ever got into the war. 

DR. E. W. KELLOGG: Since your organization pays for itself and is not entirely 
altruistic in its operation, how do you carry out the policy fostering the distribu- 
tion abroad of films that will do us the most good in the sense of promoting inter- 
national good will? 

MR. HICKS: I presume that you are referring to the Motion Picture Export 
Association which operates in the thirteen countries I mentioned. The way that 
the corporation is set up is that the eight major companies own part interest in it. 
The interest owned is hi direct relation to our normal percentage of business around 
the world. That is the same basis upon which the major producers have always 
supported the Hayes Office, and now the Johnston Office. If, for instance, Metro's 
share of the world business were 22 per cent, we would pay 22 per cent of the bills 
of the corporation and we would take 22 per cent of the income. Under those cir- 
cumstances, it doesn't make any difference to MGM whether we have any pic- 
tures hi there or not. The directors of the Motion Picture Export Association 
are entirely free to choose those pictures which they think will best depict America 

306 STOTT Vol 49, No. 4 

abroad, although they invariably consult with the Committee on Selectivity of 
M.P.A.'s International Division. 

MR. KRUSE : The point that Mr. Golden made about countries wanting to buy 
pictures attacking our way of life would have been good if it was not for the fact 
that the pictures which Russia picked were Deanna Durbin's. If you can find 
anything critical of the American way of life in that type of picture, I would like 
to see it. 

Do you not think you are mixing up propaganda and entertainment? For every 
entertainment picture that is made that is critical there are a hundred made that 
are not critical. Mr. Hicks pointed out that the Norwegians certainly resented 
this censorship. 

When the picture "Boomerang" was shown in Britain, every single critic 
praised the film because at last the Americans were getting grown up enough to 
give them pictures that were not made up of milk toast and honey. 

MR. HICKS: Let me say, Mr. Kruse, that I think you are reading into this a 
censorship which does not exist as viciously as you imply. There are about 1200 
films in our backlog pictures that were not released in those countries during the 
war. Is it not to the industry's interest as well as to America's interest to pick the 
best of those 1200 in view of the fact that we can only distribute 52 to 104 pictures 



Summary. In the commercial production of two-color prints on duplitized posi- 
tive film printed from bipack negatives, usual processing methods involve one or more 
flotation or mechanical application operations in order to prevent the treatment of the 
image on one side of the film with the color intended for the opposite side. These 
operations may be slow, difficult to control, and involve considerable waste. A method 
is outlined whereby a protective coating is applied to one side of the film so that the 
opposite side may be treated by total immersion to form the proper color for the un- 
protected image. The coating is then removed, and the previously protected side of the 
film is treated. The entire process may be done in conventional processing machines 
by total immersion of the film in the processing solutions. 

During the past two years considerable interest has been revived 
in the production of two-color prints for general 35-mm theater re- 
lease. This has been a result of mounting theatergoer demand for 

* Presented Apr. 22, 1947, at the SMPE Convention in Chicago. 
** Eastman Kodak Co., Motion Picture Film Department, East Coast Division, 
New York, N. Y. 

Oct. 1947 



motion pictures in color and the immediate lack of sufficient pro- 
duction capacity in three-color motion pictures to satisfy this de- 
mand. In addition, the production of two-color pictures is a rela- 
tively simple process as compared to three-color processes. This is 
especially true at the camera stage of the process since bipack nega- 
tives may be exposed in conventional black-and-white cameras with 
only minor modifications. 1 The printing of duplitized film from bi- 
pack negatives presents several problems not usually encountered in 
black-and-white printing since registration of the two images on 
opposite sides of the film is of prime importance, but this difficulty has 
been successfully worked out in several types of step printers. This 
paper is not concerned with these problems but rather with the prob- 
lems involved in processing the exposed duplitized film so that the 
proper color is applied to the proper image. 

A typical scheme for processing the exposed duplitized positive 
film in order to produce a two-color print is shown in the following 
outline : 

Process Step 

Black-and-White Processing 

(a) Develop in black-and-white developer ) 

(&) Stop and fix in hypo 
(c) Wash 


Silver on both sides of film 


(a) Treat one side of film in iodizing solution 

(b) Clear in bisulfite solution 

(c) Wash 

(d) Dye silver-iodide image 

(e) Wash or backwash in acid 

Prussian-Blue Toning 
(a) Immerse film in Prussian-blue toning solution ; 


(a) Wash 

(b) Fix and harden 

(c) Wash 

Dyed silver iodide on one 
side of film and silver on 
other side 

Dyed silver iodide on one 
side of film and Prussian 
blue on other side 

Silver iodide removed as 
well as silver f errocyanide 
in Prussian-blue image 
making images trans- 

As outlined above the first three steps may be done in a conven- 
tional black-and-white processing machine with undercut film spools. 

308 STOTT Voi 49, No. 4 

In the entire process the only stage that need be dark is the black- 
and-white development and sufficient treatment in the hypo or stop 
bath to arrest development. 

Numerous papers have been published and many patents granted 
which describe methods of converting a silver image on only one 
side of duplitized film to silver iodide. Various methods for applying 
a solution to only one side of the duplitized film have been described 
by Kelley, 2 Brewster, 3 Capstaff, 4 Mason, 5 Troland, Ball, and An- 
drews, 6 and others. Miller 7 and Cory 8 have described methods of 
converting a silver image in a photographic film to silver iodide suit- 
able for mordanting a basic dye. 

These mechanical .methods of iodizing one side of duplitized film 
are effective when properly handled, but considerable care must be 
exercised to prevent print damage from accidental treatment of the 
opposite side of the film. In some cases these methods may require 
slow film movement throughout the process and thus seriously limit 
production capacity for a given plant area. 

Therefore, at this stage of the process one side of the duplitized 
film has been suitably prepared such that by immersion in a basic 
dye solution the silver iodide will function as a mordant to cause 
deposition of the dye in proportion to the density of the original silver 
image. Before this treatment, however, it is necessary to remove the 
excess iodine left in the film as a result of the iodizing treatment. 
This has been described by Wall 9 and involves treating the film in a 
dilute solution of sodium bisulfite. 

After a wash to remove reaction products from the previous treat- 
ments, the film is immersed in a solution of basic dye. This dye will 
mordant only to the silver-iodide image without affecting the silver 
image on the opposite side of the film. Thus this operation may be 
done by total immersion of the film in the dye solution. Choice of 
various dye mixtures or single dyes as a satisfactory colorant for the 
print made from the bipack orthochromatic negative depends on 
many factors beyond the province of this paper. 

The film is then washed thoroughly in water to remove the excess 
dye in the gelatine of the film. Miller 7 and Cory 8 have described 
methods of backwashing the film with solutions of weak acids to 
hasten the removal of excess dye and give stain-free high lights in 
the final print. 

Crabtree and Matthews 10 have published a toning formula which 
will convert a silver image to a Prussian-blue image. The film may 


be immersed in the toning solution since the treatment has little 
effect on the previously applied dye-mordanted image on the opposite 
side of the film. Therefore, the film at this stage of the process con- 
sists of an orange-red dye-morda"nted image on one side of the film and 
a Prussian-blue image on the opposite side of the film. 

After suitable washing, the film may be fixed and hardened as 
described by Miller 7 and Cory 8 in order to render the images more 
transparent and suitable for projection. Final washing and drying 
complete the process. 

In recent years, several processes have been patented which are 
designed to improve the transparency and definition of the dye-mor- 
danted and toned images, extend the uses of duplitized film to three- 
color processes, and improve the color rendition of two-color proc- 
esses by the better choice of colorants for the two records. In 
general the chemistry of these improvements will not alter the basic 
requirements for the process herein outlined. 

It can be seen that the processing of duplitized film to yield two- 
color pictures could be made a relatively rapid and easily controlled 
system if some method could be devised to eliminate the flotation or 
mechanical application step and treat the film in conventional deep- 
tank processing machines. This has been done by applying to one 
side of the duplitized film a protective coating which is impervious 
to the treating solutions and which can be removed easily without 
altering the characteristics of the treatment applied to the unpro- 
tected side of the film. The idea of Using ' 'resists' ' for various types of 
photographic processing is not new having been described by many 
inventors including Kelley, 2 Shorrocks, 11 Mannes and Godowsky, 12 
and Lierg. 13 Using a protective coating which is impervious to an 
iodizing solution, the following scheme could be used to process du- 
plitized film by total immersion of the film in the processing solutions 
at every stage of the process. 

It can be seen that the rate of film movement through the process 
is limited only by the usual limitations of a continuous processing 
machine with a multiplicity of tanks. Control of the iodizing step 
is simplified since the solution may be vigorously agitated with no 
danger of accidental treatment of the opposite side of the film. 

The key to this process is the protective coating applied to one 
side of the duplitized film. Several types of -protective ' coatings 1 
were considered. Tests were made on materials removable from the 
film by treatment in solutions of strong acids. It ds evident -4hat 



Vol 49, No. 4 

selection of such a material for this specific purpose would be difficult 
since the iodizing solution functions most effectively when acid. It 
was decided that this type of protective coating was not practicable 
for this type of process. . 

Process Step 

Black-and-White Processing 


(a) Develop in black-and-white developer 

(6) Stop and fix in hypo 

(c) Wash 

(<*) Dry 

(e) Apply protective coating to one side of film 

(/) Dry 

Silver on both sides of film 




Iodize in deep tank 

Clear in bisulfite solution 

Remove protective coating 


Dye silver-iodide image 

Wash or backwash in acid 

Prussian-Blue Toning 

(a) Immerse film in Prussian-blue toning solution 


(a) Wash 

(b) Fix and harden 

(c) Wash 
C<9 Dry 

Dyed silver iodide on one 
side of film arid protected 
silver image on opposite 

Dyed silver iodide on one 
side of film and Prussian- 
blue image on opposite 

Silver iodide removed as 
well as silver ferrocyanide 
in Prussian-blue image 
making final images trans- 

Attention was then turned to protective coatings that are soluble 
in alkaline solution. If such a material could be found that would 
exclude the iodizing solution from the emulsion of the protected 
side of the film, the system should be simple to control since, as men- 
tioned before, the iodizing solution functions most effectively when 
acid. Thus the coating would remain intact throughout the iodizing 
treatment and be removable in an alkaline solution which would not 
affect the iodized image. 

One alkali-soluble coating material which is familiar to the motion 
picture industry is Eastman Universal Protective Film Lacquer. 


This product was originally described by Talbot 14 in connection with 
the protection of finished prints from scratching, abrasion, and oil- 
mottling during projection. In designing this product it was 
known that it is difficult if not impossible to prevent abrasion and 
scratching of projected prints with any type of protective coating. 
Therefore, it was reasoned that with a protective coating on the film, 
the majority of the scratches and abrasions suffered during projec- 
tion would be confined to the thickness of the coating. The lacquer 

FIG. 1. Apparatus for "bead" application of lacquer 
to film. 

was so formulated that when the coated film is treated in a dilute 
solution of sodium carbonate or in a standard positive developer 
followed by a short water wash, the lacquer is removed without 
buffing or scrubbing. Thus the scratches and abrasions will disap- 
pear with the lacquer. 

Tests were made on this lacquer. These original tests were re- 
markably successful although it was immediately learned that a 
somewhat different lacquer formula was required. It was also 
learned that a heavier coating than ordinarily is necessary for print 
protection during projection was advisable. The most satisfactory 

312 STOTT Vol 49, No. 4 

results are obtained when the lacquer is "-bead-applied" since this 
gives greater control of the coating thickness over a wide range of 
machine speeds. Whereas a coating thickness of about 0.0001 inch 
is usually satisfactory for adequate mechanical print protection, it is 
advisable to have a coating thickness of about 0.0002 inch for chemi- 
cal protection of one side of the film during this type of process- 
ing. Using this lacquer spread more than 10,000 feet of 35-mm film 
may be coated on one side with one gallon of lacquer. 

The lacquer must be applied to dry film. Therefore, the bead 
applicator may be installed in the drying cabinet of the black-and- 
white stage of the process such that several drying loops are available 
after the application of the lacquer to dry the coating properly. 

Fig. 1 shows a bead-application device which has been used suc- 
cessfully in this type of work. In this device, as the film passes over 
an idler roller, lacquer is applied at the bottom surface from a second 
applicator roller spaced just out of contact with the film. This ap- 
plicator roller is slowly driven and carries the lacquer up from a pan 
(not shown) and applies it to the film through a liquid bead main- 
tained by the surface tension of the lacquer. 

Studies were made on how much undercutting could be expected 
at the edge of the lacquer coating by the iodizing treatment. It is 
interesting to note that for the iodizing treatments used in these 
studies this undercutting did not exceed 0.001 inch. However, this 
figure will vary depending upon the time of the treatments and the 
iodizing solution used. 

In "bead-application" of a lacquer, the coating cannot be applied 
over the entire area of the film since the lacquer will wet through the 
perforations and cause transfer on the opposite side of the film due 
to wetting of the rollers of the application device. In practice, it 
has been found that the lacquer coating may be bead-applied well 
into the perforation area without causing leakage of the lacquer 
through the perforations and thus transfer on the opposite side of 
the film. This wider coating provides a margin of safety several 
times greater than the amount of undercutting to be expected, and 
thus insures against contamination in the sound track or picture area 
due to this effect. 

Fig. 2 illustrates the above point. Here a length of Eastman 35- 
mm Duplitized Positive Film has been flash-exposed across the entire 
film area on only one side of the film. This film was then processed 
to form, a black-and-white image, lacquer-coated using the bead 


method, and then iodized, cleared, fixed, washed, and dried. Thus 
the black pattern within the clear perforation area represents the 
area of the film protected by the lacquer coating. It is apparent 
that any slight undercutting of the lacquer coating occurring here 
will not affect either the picture or sound track. 

It was found in studying this process that careful application of the 
lacquer to the film eliminates many difficulties in the subsequent 
processing. Some of the early tests revealed minute pinholes in the 
coating which caused contamination of the protected side of the film 
with tiny specks of the wrong color. It was learned that these were 

FIG. 2. Darkened portion of film indicates area to which 
lacquer is applied by bead applicator. 

caused by minute particles of dirt that had deposited on the film 
either during wet processing, drying, or handling. Careful filtering 
of the processing solutions, and the air used in drying the film, and 
care in handling eliminated this difficulty. 

For various reasons it may be desirable to alter the order of treat- 
ment in making two-color prints. Since the protective coating is 
also impervious to the Prussian-blue toning solution, the image 
printed from the ortho negative may be protected with the lacquer 
coating and the Prussian-blue tone applied to the unprotected pan 
negative print first. Should this be done, a minor complication 
arises. The subsequent treatment in alkaline solution to remove the 
lacquer will render the Prussian-blue image brown. This is caused 

314 STOTT Vol 49, No. 4 

by the fact that Prussian blue (ferric ferrocyanide) is converted by 
alkali to ferric hydroxide which has a characteristic brown color. 
Ferric hydroxide is quite insoluble in alkaline solution so the image 
does not dissolve completely but redeposits in the film. Thus after 
the alkaline treatment to remove the lacquer, the brown image may 
be converted to the original Prussian-blue image by treatment in an 
acid solution of ferrocyanide. Inversion of the process in this fashion 
thus adds one more chemical processing step and one more wash. 
It is also possible to work out other processes such that the red image 
is formed by toning and the blue image by dye-mordanting. Like- 
wise both images may be formed by dye-mordanting by the use of 
two "resists" at various stages of the process. This procedure may 
be desirable in order to provide better colorants for the two records 
and thus improve final screen results. 

This type of alkali soluble "resist" may be used for these processes 
only when the protective treating solution is acid. It has been found 
that this lacquer coating when properly applied will provide ade- 
quate protection to the coated emulsion as long as the treating solu- 
tion has a pB. less than 3.5. However, where it is possible to choose 
between a dilute treating solution and a longer treating time and a 
more concentrated solution and a shorter treating time, in most 
cases it would be advisable to choose the latter procedure in order to 
minimize the possibility of penetration or deterioration of the coating 
even in the acid solution. These details would have to be worked 
out for the particular process being developed. 

The lacquer coating is removed by immersion of the film in a 2 
per cent solution of sodium carbonate for two minutes followed by a 
two-minute wash in running water. Although the lacquer does not 
dissolve completely in the removal solution but tends to loosen and 
slip off the film, the removal solution may be replenished in the usual 
manner as long as the rate of replenishment is high enough to pre- 
vent sludge accumulation. 

No attempt has been made to produce any length of two-color film 
suitable for projection using this process. All of the work has been 
done on a small scale, and most of the studies have been made on the 
investigation of the protective aspects of the lacquer coating only. 
The chemistry of this type of process is in general straightforward and 
is covered thoroughly in the literature. Although it is appreciated 
that testing of this protective lacquer for this use has not been ex- 
tended to actual machine processing, it is felt that these studies have 


indicated that the product will function satisfactorily in practice for 
the purpose herein outlined. 


The author wishes to express his appreciation to Mr. R. H. Talbot 
for many helpful suggestions and his work in compounding the new lac- 
quer and preparing the film samples on which these studies were made. 


1 BOYLE, J. W., AND BERG, B.: J. Soc. Mot. Pict. Eng., 48, 2 (Feb. 1947), pp. 

2 KELLEY, W. V. D. : United States Patent 1,259,411 ; French Patent 539,936. 

3 BREWSTER, P. D.: United States Patent 1,223,664 (1917). 

4 CAPSTAFF, J. G.: United States Patent 1,351,834 (1920). 

6 MASON, J.: United States Patent 1,348,029; Deutsche Reich Patentschriften 
344,352; French Patent 517,577; abst. Sci. Tech. Ind. Phot., 1 (1921), p. 80; 
British Patent 143,230; abst. /. Soc. Chem. Ind., 40 (1921), p. 530A; Eder's 
Jahrbuch, 29 (1915-1920), p. 159. 

6 TROLAND, L. T., BALL, J. A., AND ANDREWS, J. M.: United States Patent 
1,435,764 (1922) ; French Patent 558,386; abst. Sci. Ind. Phot., 4 (1924), p. 31; 
5 (1925), p. 49; British Patent 211,918; Brit. J. Phot., 71 (1924), p. 418; Deutsche 
Reich Patentschriften 395,547. 

7 MILLER, H.: Brit. J. Phot., 54 (1907), pp. 136, 196, 216; Phot. Korr., 44 
(1907), p. 55; Phot. Chron., 14 (1907), p. 207. 

8 CORY, A. S.: Mot. Pict: News, 15 (June 2, 1917), p. 3488. 

9 WALL, E. J. : "History of Three Color Photography," American Photographic 
Publishing Co., (Boston, Mass.), 1925. 

10 CRABTREE, J. I., AND MATTHEWS, G. E.: "Photographic Chemicals and Solu- 
tions," American Photographic Publishing Co., (Boston, Mass.,) 1939. 

11 SHORROCKS, H.: United States Patent 1,303,506; British Patent 111,054. 

12 MANNES, L. D., AND GODOWSKY, L. : United States Patent 2,010,459. 

13 LIERG, F.: British Patent 333,697. 

14 TALBOT, R. H.: /. Soc. Mot. Pict. Eng., 37, 2 (Feb. 1941), pp. 191-197. 


CHAIRMAN FILLIMORE : Is this process in use commercially or is it just in the 
laboratory stages at this moment? 

MR. J. G. STOTT: It has not been commercially applied. 

CHAIRMAN FILLIMORE: Will it be commercially available in a short time? 

MR. STOTT: The lacquer is available immediately. The chief thing that is 
required is to set up a bead-application device hi the drying cabinet of the black- 
and-white stage of the processing machine. Adjustment should be made such that 
the coating thickness is sufficient. 

MEMBER: What is the name of the lacquer? 

MR. STOTT: Eastman Protective Film Lacquer. 

CHAIRMAN FILLIMORE: What is the iodizing solution? 

MR. STOTT : There are a large variety of iodizing solutions. One type is a solu- 
tion of potassium iodate, potassium iodide, and acetic acid. 



Summary. The duplicating process has been studied at five Hollywood labora- 
tories. Prints made both by the duplicating process and directly from an original 
negative have been obtained from all five laboratories. No attempt was made in this 
study to include the sound track. 

A complete sensitometric study was made using both the printed-through technique 
and customary Type lib Sensitometer gamma controls for each step in the duplicating 
process. Resolving-power measurements throughout the various steps in the process 
were also obtained. 

Considerable variation has been found in the degree of fidelity of the duplicating 
procedure at the five laboratories studied. In most instances it is possible to determine 
the cause or causes of poor quality by examination of the sensitometric data. Under- 
exposure of either the master positive or the duplicate negative or both is found to be 
the greatest cause of lack of proper tone reproduction in the duplicate prints. 


Since the inception of fine-grain duplicating films in 1936 1 very 
little has been written about the motion picture duplicating process. 
The use of duplicating films has increased disproportionately in 
comparison with the use of motion picture films in general. Com- 
posite duplicate negatives or master positives are sent abroad for 
foreign release printing. In many instances the final edited negative 
for domestic release printing consists of 50 per cent or more duplicate 
negative. When such a release negative is duplicated in toto in order 
to provide foreign release negatives, a corresponding portion of the 
latter becomes a second-generation duplicate. 

For all subject matter requiring the use of an optical printer, 
duplicating becomes a necessity. The optical printing technique 
has- acquired great significance. It is used for making lap dissolves, 
wipes, blowups of a portion of the original scene, straightening an 
exposure which was originally shot with a tilted camera, and so on. 

Less frequently the duplicating process has been utilized to cor- 
rect mistakes made in the exposure or development of the original 
negative. An overexposed negative, which prints outside the normal 

* Presented Apr. 22, 1947, at the SMPE Convention in Chicago. 
** Eastman Kodak Company, Motion Picture Film Department, West Coast 
Division, Hollywood, Calif. 


printing light scale, may be brought down to a reasonable printing 
light level by making a duplicate negative. Errors in contrast of 
the original negative may be easily corrected by adjusting the de- 
velopment of a duplicate negative. It is not possible, of course, to 
correct for too great a degree of underexposure in the original nega- 
tive by making a duplicate negative. However, if an original nega- 
tive is somewhat underexposed and suffers from low contrast because 
of this underexposure, this condition may be palliated somewhat by 
making a correctly exposed duplicate negative at higher than normal 

In order to determine the degree of fidelity in reproduction of the 
picture negative by the duplicating process, periodic survey tests of 
the major laboratories in Hollywood have been made by the authors. 
The results of the most recent of these tests will be presented here. 


A representative picture negative photographed on Plus X film 
and developed to a lib gamma of 0.65, was chosen as the starting 
point for all tests. This girl-head negative had high-light and shadow 
densities of 1.48 and 0.41. These densities were read on a Western 
Electric RA-1100B densitometer, with blue printing filter, using 
a specially designed circular aperture, 13 mils in diameter. This 
small aperture has proved most helpful in scanning actual picture 
scenes. High-light and shadow densities reported for all films were 
read in this manner. With a little practice it is possible to make such 
readings in the picture area with a repeat accuracy of 0.01 den- 
sity unit. Attached to a 100-foot length of the picture sample was a 
10-foot length of resolving-power test negative in which the maximum 
resolution test chart revealed 56 lines per millimeter with sharp 
detail. Also, a specially prepared full-frame step-tablet was at- 
tached, the steps being increments of 0.15 density. The tablet con- 
tained 15 such steps. 

This standard negative was submitted to each of five laboratories 
in turn and each laboratory was requested to make a master positive, 
a duplicate negative, a print from the duplicate negative, and a print 
from the original negative. It was pointed out to each laboratory 
superintendent that the purpose of the test was to match the two 
prints as closely as possible. 

The original negative was not developed to the exact degree of 
contrast normally used by some of these five laboratories. As a 

318 SIMMONS AND HUSE Vol49, No. 4 

Q 2 Q 

1 1 

* ; a |i ? I a |i j 3 g ji |.__ 

13 13 15 

O bb O bo O 

Q r{ ^S'2 * * S'2 a * 

ffi - g^ft g._ co^cx g fe ffi - 

_K) O rt fy. 

OQ o ^^ J2 ^^ 


O H-l 10 -a "O H PH 

13 13 13 

1 i * I i i- 1 

O^ go,C<5 C<5 

" -^^^ 4> ^ ''-'a V HrT ^Q 

dy .o^c'^ -2 d8 C fe ^w d^ woCfo 

PQ -^ io^O H CO *O O5 Pk Tf 1 PQ CQ^O H PH S t> PH O PQcO^oHP^ 

Q W 

g bib 'o "8 

d'o 1 fi S C'o* a ^ S3 "5* 


^O H Pin' N ^ PH W 

^* -* ^"^ fi ^5 Cb' fi ^ C'O 

S <K . i-^ 5Pr." i* ^H Z^ r*-* ^ ^^ i r-H 2?rv* 

S^o d 

I | * J * f 

M _.^,<L) >^ .^,(U 

<^ g2 c ? S2 s o 

J llrio is 5 |f is ^ 


j| g g> I -5 .5 J f I & ' 311 

i 2 fc.'J 1 1 8 & 9 1 o^*^ IH "2 tJ oo o-cxa g.-3 w, "g tj' o o o. a o, 

>J u-Saftx'&rtaJoj'S^ SSft^^^rtrt"'"^^ 'Sa^><XrtcJrt'Pi 

5 .2 ^ <j w w .-} ij 1-3 Q aJ^<ww4i2Ji3 ^o ^.Sw<:ww 
3 Q ^ Q .Sfi 

S Q fi 




!S -8SSIfeS 3 SSSSS gSSg 

Co---- CM---- CYJ . . .. . .. 


-H<M CM 

*-n OO TH o H i 

GO CO * iO O O TH CM CO 00 




Co TH CO lO, 

TH" TH TH 0* 






H 00 TH CO 
B "* 

TH TH o 



H O O TH o" TH CM* CM O CM* CM" O <M" 

W " H 


CM - 

*H TH O 




1 TH Tj* CO CM 






- g I ^ 'bb fe .5 

T-^-T b ^-^ V-T l> 

rtg^Trjo g b "e|' c !'S 



320 SIMMONS AND HUSE Vol 49, No. 4 

result, a comparison of the final over-all screen contrast at the five 
different laboratories has little significance. The variations in the 
screen contrast of the final prints are a measure of the actual printing 
variation rather than a true measure of the normal over-all produc- 
tion contrast for each laboratory. In order to compare over-all 
screen contrast of each laboratory's product it would be necessary to 
have each laboratory develop its own negative and then print that, 
rather than a standard test negative. The comparison of the print 
made from the duplicate negative with the print made from the 
original negative is completely valid, however, for each of the five 

The manner of reading all densities reported in this paper is that 
which has been in use for some time at this laboratory. The den- 
sities of those processed films which were printed onto blue-sensitive, 
or positive-type raw film, were read with diffuse blue light. In the 
case of films which were to be printed onto panchromatic film, visual 
diffuse densities were determined. Frayne 2 has promulgated this 
technique of reading densities in the manner equivalent to the print- 
ing conditions used in his excellent paper on the measurement of 
photographic printing density. 

The resolving-power measurements were made by reading at five 
positions in each of several frames and using the maximum figure 
found in each individual frame for averaging purposes. Thus where 
resolution was not equally good over the entire picture area, the data 
reported represent the best values, rather than the over-all picture- 
area average. In a number of instances resolution was better in 
one half of the picture area than in the other half. 

In Tables 1 and 2 are given the printing and developing data and 
the sensitometric results of the tests. 


Eastman Fine Grain Duplicating Positive, Type 1365, was used 
for making the master positive by each of the five laboratories where 
tests were made. The gamma value to which this film is developed 
varies considerably. Fig. 1 shows the lib sensitometric exposure 
curves and lists the manner in which this film is exposed on the 
type lib sensitometer at each laboratory. Since this method of ex- 
posure is useful only for control purposes, it is of little importance 
whether the positive conversion filter is left in place or removed from 
the sensitometer in order to obtain more exposure. It is essential, 



however, that sufficient exposure be given to 1365, by multiple ex- 
posure and/or by removing the conversion filter, so that a reasonable 
number of the density steps obtained lie on the straight-line portion 
of the H and D curve. 

In order to speed up the making of routine daily control strips, it 
has been recommended that the color temperature conversion filter 
be removed from the lib sensitometer and 1365 be given four ex- 
posures. If the filter is left in place it is considered good practice to 
use eight exposures. 

3 6 
3 2 









FIG. 1. lib sensitometric exposure curves for Eastman duplicating positive 

film, type 1365. 

The lib gamma values for 1365 vary from 1.12 to 1.54 at the five 
laboratories. The eleventh step density on the lib step-tablet has 
been indicated on each curve by a short crossbar. This designation 
has been used for the reference step on all other lib exposure curves 
in this paper. 

In Fig. 2 are shown the print-through curves for the 1365 master 
positives. It will be seen that the print- through gamma values are 
somewhat higher than the corresponding lib control gamma values. 
The lib and print-through exposure strips were, of course, developed 
together with the picture sample at each laboratory. 

The seventh step density in the 15-frame standard step-tablet has 
been marked on each curve in Fig. 2 so that these print-through 



Vol 49, No. 4 

curves will be more useful for comparison purposes. This procedure 
has been continued in the subsequent print-through data given in 
this paper. The density of this step in the original negative tablet 
was 1.03. 

In order to satisfy the primary requirement for good tone repro- 
duction in the duplicating process, it is essential that all densities in 
the master positive lie on the straight-line portion of the print- 
through H and D curve. By examining the master positive high- 
light densities in Table 2 and by placing them on the print-through 


FIG. 2. Printer exposure curves for Eastman duplicating positive film, 

type 1365. 

curves in Fig. 2, it may be seen that in some instances there is definite 
underexposure of the master positive. Furthermore, by placing the 
shadow densities from Table 2 on the print-through curves it may be 
seen that considerably higher exposure might have been used in all 
instances without causing any loss of shadow detail due to the higher 
densities falling on the H and D curve shoulder. In a large number of 
such sensitometri6 analyses of master positives made over several 
years, there has not been noted a single instance where densities 
occur in the scene higher than the shoulder break density of 1365. 

Most Hollywood laboratories examine the master positive by pro- 
jection, using special projection facilities in order to avoid damaging 
the film in any way. It is apparent that an exceedingly dark master 


positive is most difficult to examine in this manner for minor defects, 
such as dirt. Consequently it is the natural inclination of the 
laboratory to print the master positive lighter, so that it can be 
judged more critically on the screen. This results in a high-light 
tone distortion due to the use of the H and D curve toe. It is be- 
lieved that the benefit to be gained by many laboratories from print- 
ing the master positive considerably darker than is being done at 
present, and not using projection examination, would more than off- 
set the quality loss which may result from lack of rigid screen in- 
spection of the master positive. 

The low emulsion speed of 1365, and lack of sufficient printer light 
to expose this film properly, is a second cause of underexposure of the 
master positive. Where full advantage of the recommended method 3 
of increasing printer illumination efficiency has been taken, there is 
no difficulty in obtaining proper exposure on 1 365. 

It is apparent from Fig. 2 that a lower high-light density is allow- 
able in those instances where lower gamma is used. As so often 
noted in Hollywood, the master positives from those laboratories 
which make the best duplicate prints are not necessarily the darkest 
master positives. This is true because of the wide variation in gamma 
used for the master positive and the consequent variation in toe- 
break density. 

Equally good master positives may be made on 1365 at print- 
through gamma values ranging from 1.00 to 2.00. It is only neces- 
sary that the duplicate negative gamma value be balanced accord- 
ingly. The minimum allowable exposure necessary to place all 
densities in the master positive on the straight-line portion of the 
curve remains quite constant as the degree of development increases. 
Thus the effective emulsion speed is not increased by working at a 
higher gamma value, as is often erroneously thought to be the case for 
this film. 

All laboratories make printer light changes from scene to scene in 
printing the master positive. If timing of a negative is first done in 
the making of a release print, it should be possible to use exactly the 
same light changes, with a much increased exposure source, in making 
the master positive. The lower gamma of the master positive will 
result in a lower density increment per printer light, but this com- 
pression of the density scale is, of course, as it should be. The print- 
ing scale for 1365 should never be adjusted so that the density incre- 
ment per printer light becomes the same as that obtained on 1302. 

324 SIMMONS AND HUSE Vol 49, No. 4 

If this is done an abnormal increase of density in the darker scenes 
and lessening of density in the lighter scenes will be evident in the 
final print. Considerable effort is expended by some laboratories in 
selectively timing the master positive; sometimes light changes are 
made between scenes that do not call for light changes in the release 
print timing. This is inconsistent and is certain to result in lack of 
exact duplication of the print made directly from the original nega- 
tive, unless a compensating light change is made in printing the du- 
plicate negative or the print from the duplicate negative. 

Since almost all laboratories strive to make the duplicate negative 
and the final duplicate print without making printer light changes, 
the production of a correctly timed master positive is of great 

Many laboratories have the master positive printer exposure scale 
and over-all illumination so adjusted that the printer lights used for 
the release print are automatically correct for the master positive 
printer. Of course, the same printer setup cannot be used for both 
jobs. By properly attenuating the illumination without affecting 
the relative exposure scale, the same printer that is set up for printing 
master positives may be used for making release prints. Obviously 
if the lightest scene in a master positive is printed dark enough to 
place the high lights on the straight-line portion of the curve, then 
some scenes will be considerably darker than this minimum allow- 
able high-light density. This is necessary if a "one-light" duplicate 
negative is to be made. 

The resolving-power data in Table 2 show that 56 lines per milli- 
meter, which was the maximum resolution chart used in the original 
negative, may still be resolved in all five master positives. There is 
undoubtedly some loss in resolution but it is less than could be 
measured by the test charts used. 


About two years ago a change was made in the emulsion charac- 
teristics of Eastman Fine-Grain Panchromatic Duplicating Nega- 
tive Film, Type 1203, which was designed to accomplish three aims. 

1. To increase the latitude, or length of avilable straight-line 
portion of the H and D curve. 

2. To increase the developing time necessary for normally used 
gamma values. This was desirable in view of the abnormally short 
developing times which were dictated by economic considerations at 
some laboratories. 



3 . To give better uniform ity and freedom from flicker, unevenness, 
etc., in the screen image of duplicate prints. 

Figs. 3 and 4 show the lib and print- through H and D curves for 
1203 Sit five laboratories in Hollywood. A single normal positive 
exposure was used on the lib sensitometer in all cases. The print- 
through curves have generally higher gamma values. In comparing 
the contrasts of the duplicate negative at the several laboratories it 
is apparent that the lib gamma values are misleading. This ac- 
counts for much confusion in the minds of those laboratory men who 




FIG. 3. lib sensitometric exposure curves for Eastman fine-grain pan- 
chromatic duplicating negative film, type 1203. 

are well aware of the large differences in contrast in the master posi- 
tives from various laboratories and who cannot understand how this 
has been compensated for in the duplicate negative when they com- 
pare the lib gamma values of the latter. 

If the shapes of the curves in Fig. 3 are compared with the similar 
lib exposure curves for 1203 which were published by Ives and Crab- 
tree 1 in September, 1937, the increase in latitude in the current film 
will be readily seen. 

In order that the requirements for correct tone reproduction be 
satisfied, it is essential that the full scale of densities in the original 
negative be placed entirely on the straight-line portion of the dupli- 
cate negative. The effect of deviations from this manner of exposure 



Vol 49, No. 4 

for the duplicate negative may be obtained by the use of a graphical 
method for the study of tone reproduction described by Jones. 4 

A large portion of all original motion picture camera negative is 
exposed so that the toe of the H and D curve is utilized, and, in fact, 
so that there are generally shadow areas to be found, which, upon 
measurement, show no silver density whatsoever. In order that a 
satisfactory duplicate negative be made of such an original negative, 
the duplicate negative must be so exposed that the minimum density 
in the original negative, which was base density, shall become the 
toe-break density, or somewhat higher, in the duplicate negative. 


id 0.8 



FIG. 4. Printer exposure curves for Eastman fine-grain panchromatic dupli- 
cating negative film, type 1203. 

Obviously, therefore, the duplicate negative would be darker than 
the original negative and would require a higher printing light in 
order to given an exactly matching print. This fact is well known 
to most laboratory men. However, the erroneous impression still 
exists in many places that a duplicate negative, made from even a 
badly underexposed original negative, should match the original 
negative, density for density, and, if made perfectly, should print 
on the same printing light. It is true that, because of the lower toe 
break of the duplicating negative film, compared with that of most 
picture negative materials, a correctly exposed duplicate negative 
may have lower printing density than the original negative from 
which it was made, in those instances where the original negative 


had shadow density at or about the toe break. There are many in- 
stances where the original negative, being overexposed, will be con- 
siderably heavier in density than a correctly made duplicate negative. 

As pointed out by Ives and Crab tree, 1 there is a difference in the 
ratio of visual to effective printing densities for 1203 and the ordinary 
picture negative films. This difference causes a compression of the 
density scale in the duplicate negative relative to the original nega- 
tive, if densities are read with visual light. If, however, densities 
are read with blue printing light, as described earlier in this paper, 
this discrepancy largely disappears and the actual printing behavior 
of the duplicate negative relative to the original negative may be 
more accurately predicated. 

There are special instances where the application of the classical 
methods of tone reproduction cannot be applied to the duplicating 
process. For example, in the making of ordinary dissolves, lap dis- 
solves, and other such special effects, oftentimes the entire scene is 
not duplicated. If only a portion of the scene which is to contain the 
special effect is duplicated, it is necessary that there be no printer 
light change at the point of juncture of the original negative and the 
duplicate negative. If there is a printer light change at this point, 
it is almost surely to be noticeable as a "bounce" in illumination on 
the screen. Consequently the printing densities of the duplicate and 
original negatives must be as nearly equal as possible. This implies 
that if the original negative were badly underexposed it is necessary 
to make an underexposed duplicate negative to match it. The re- 
sulting superimposition of the toe region of the duplicate negative 
curve on the toe of the original negative results in serious loss of con- 
trast in the shadow areas. 

In the making of certain trick shots, where black velvet covers 
large areas of the scene being photographed, and double exposure 
is used in making a composite duplicate negative, usually in an op- 
tical printer, it is necessary that there be no printed-in density from 
the black velvet in the duplicate negative. Therefore, one must ex- 
pose so as to obtain clear film base in the duplicate negative in those 
areas covered by black velvet in the original scene. This procedure 
makes it very difficult to obtain linear tone reproduction in the du- 
plicate negative for the points of interest in the scene. However, by 
proper lighting, this technique, used so successfully in such pictures 
as "The Invisible Man" and many others, both in black and white 
and color, is accomplished with very little degradation in quality. 



Vol 49, No. 4 

If the high-light and shadow densities of the duplicate negatives 
listed in Table 2 are fitted onto the print-through curves in Fig. 4, 
it will be seen that, just as in the case of the master positive, there is a 
general tendency to underexpose the duplicate negative. This re- 
sults in a loss of shadow detail in the duplicate print. It is believed 
that this general tendency to underexpose the duplicate negative 
stems partially from the feeling that the duplicate negative must be 
no darker (as judged by printing behavior) than the original negative. 


FIG. 5. lib sensitometric exposure curves for Eastman fine-grain release 
positive film, type 1302. 

The resolving-power data in Table 2 indicate a considerable 
loss in resolution in making the duplicate negatives from the master 
positives. The maximum resolving power of 1203 emulsion is ap- 
proximately 150 lines per millimeter when processed in a normal 
manner. The loss cannot be explained satisfactorily therefore in 
terms of emulsion limitations. At the present time the cause or 
causes of this loss of resolution in contact printing of duplicate nega- 
tives is not known. 

The lib sensitometric control curves for the 1302 prints made at 
the five laboratories are shown in Fig. 5. The print- through gamma 
values for the prints from original negatives and the prints from 
duplicate negatives are listed in Table 2. For the sake of brevity, 
the prints shall be referred to as "original" and "duplicate" prints. 



Since the print-through curves consist of the print densities plotted 
versus the effective printing densities of the original and duplicate 
negatives, respectively, one would expect to find no difference in 
the gamma values of the two print-through curves. However, the 
print-through gamma of the original print is lower in some instances 
and higher in others than that of the duplicate print. We interpret 
this to mean that measurement of the negative densities with diffuse 
blue light, as done on the Western Electric RA-1100B densitometer, 



FIG. 6. Over-all sensitometric comparison of prints from original negative 
with prints from duplicate negatives. 

while being a close approximation to the true printing densities, is 
not entirely accurate. The variation in the ratio of the two print- 
through gamma values may be due in part to the known variation in 
image color obtained on the duplicate negative at the different con- 
ditions of processing. The print-through curves for^ the original 
prints are shown in Fig. 6; those of the duplicate prints are not shown 
in the figures. 

In Figs. 6 and 7 are shown two commonly used methods for meas- 
uring sensitometrically the degree of fidelity of the duplicating proc- 
ess. In Fig. 6 the densities of both original and duplicate prints 
(step-tablets) are plotted versus a fixed abscissa the densities of 
the original negative. If perfect tone reproduction is achieved 
throughout the entire print density scale, the two curves should 



Vol 49, No. 4 

exactly superimpose. Fig. 7 utilizes a manner of plotting the data 
which emphasizes the departures from perfect tone reproduction. 
Here we find the same two sets of print densities used in Fig. 6 plotted 
versus each other. In the case of perfect tone reproduction, the 
dashed line at a 45-degree angle would be obtained. The actual 
comparisons of original and duplicate prints at five Hollywood 
laboratories are indicated by the solid curves. 

An analysis of these curves shows that, in general, the over-all 
contrast comparison, indicated by the deviation of slope of the solid 
curve from the dashed line, is very good. This is the natural result 



FIG. 7. Over-all fidelity of tone reproduction. 

of much day-to-day examination of finished print comparisons and 
the making of small adjustments in duplicate negative or master 
positive contrast. It is not difficult to detect small errors in over-all 
contrast by visual examination of the finished prints. 

The fidelity of tone reproduction in the extreme high-light and 
shadow regions is not so good in all cases. Underexposure of the 
master positive results in the lack of clean high lights in the duplicate 
print. Underexposure of the duplicate negative results in loss of 
shadow detail and lack of sufficiently dense blacks in the duplicate; 
print. It is very difficult to diagnose correctly these common faults 
by screen examination of the finished prints. One is much too prone 
to call the fault an error in over-all contrast, when, in actuality, it is 


an error in exposure, in one or both steps in the process. A true error 
in over-all contrast is very easily discernible and likewise easily rem- 
edied. Sensitometric analysis of the type shown in Fig. 7 is a rapid 
aid to diagnosis of the source of the more common errors of exposure. 
The resolving-power measurements in Table 2 show that the orig- 
inal prints have retained 56 lines per millimeter in all cases but one. 
Likewise, there is little additional loss in printing from the duplicate 
negatives. The total loss of resolution in the duplicate prints as com- 
pared to the original prints is large enough so that it may be detected 
upon critical screen examination if the subject matter contains con- 
trasty fine detail. A notable example is women's jewelry, with flash- 
ing pinpoint high lights . It has been observed that this type of subject 
matter is often less sharp in duplicate print's than in original prints. 


For the most precise Sensitometric evaluation of the duplicating 
process it is necessary to consider the "adjacency" effects, which in the 
case of unidirectional travel of motion picture film in a developing 
machine are known more specifically as "directional" effects. This 
has been discussed in this JOURNAL by Weiss. 5 Special directional 
effect tablets were included in the test negative used in this work, but 
no effort has been made to incorporate those data in this paper. The 
full-frame step-tablet used for these tests, while not exactly indicative 
of the conditions found in the actual picture scene, is useful for ob- 
taining comparative results and for determining the primary causes 
of degradation in the duplicating process. 

While this paper is confined to a consideration of the picture-du- 
plicating process, it is apropos to mention that since the variable- 
density sound-recording process is based on a linear relationship be- 
tween original negative exposure and final print transmission, all 
statements regarding the proper exposure and development of the 
master positive and duplicate negative apply equally well to the du- 
plication of this type of sound track. As for the variable-area sound 
track, the best values of density in the master positive and duplicate 
negative should be determined by cross-modulation tests. If this 
cannot be done, it is advisable to make listening tests, with emphasis 
on elimination of sibilant distortion, in order to set up the optimum- 
density requirements. It is usually more convenient to establish a 
fixed density in the duplicate negative which is the highest value 
readily obtainable without accompanying printer flare, and then to 

332 SIMMONS AND HUSE Vol 49, No. 4 

vary the density of the master positive sound track until the best 
conditions are reached. 


For the establishment of correct picture exposure and develop- 
ment conditions for the master positive and duplicate negative, a 
few simple steps may be enumerated. 

1. Choose a development condition which gives a print-through 
gamma value on Fine-Grain Duplicating Positive Film, Type 1365 
in the neighborhood of 1.20 to 1.50. A satisfactory H and D curve 
with linearity comparable to those shown in this paper should be 
achieved. Establish a means of controlling the development, such as 
by the use of the lib sensitometer. 

2. Make exposure tests using typical camera negatives and de- 
termine, if possible by spot measurement in the picture area, when 
the minimum density in 1365 lies above the toe break. If this cannot 
be done, determine the exposure just necessary to produce the toe- 
break density on 1365 when printed from a density of 1.50 on the 
camera negative. This can be done by using a lib step-tablet. This 
exposure should be established as the middle of the available 1365 
printing scale. This arbitrary value of negative density is representa- 
tive of the average high-light density of a fully exposed motion pic- 
ture camera negative. 

3. Make print-through exposures on 1365 to serve as a control on 
the over-all development and printing condition. The lib sensitome- 
ter is useful as a control on the developer alone. 

4. Print a correctly exposed and developed 1365 picture scene 
onto Fine-Grain Panchromatic Duplicating Negative Film, Type 
1203 so as to obtain minimum density above the toe break. This 
may be approximated by visual comparison of the picture scene with 
known step-tablet densities. Develop for a series of times, so as to 
obtain print-through gamma values of, for example, 0.60, 0.65, 0.70, 
and 0.75. Print these negatives, and also the original negative and 
compare by projection. Determine exact duplicate negative gamma 
necessary by making further tests. 

5. Make print-through exposures and lib exposures, if possible 
on 1203 to serve as controls on the development and printing opera- 

6. Plot the final print-through data as shown in Fig. 6, and de 
termine, by comparison of picture and step-tablet densities, id 


possible, whether exposure and development conditions are correctly 



The tests reported here, as well as all others made over a period of 
several years, have shown that the lack of correct tone reproduction 
in the motion picture duplicating .process may be traced by sensi- 
tometric analysis to two primary causes : 

1. Underexposure of the master positive. 

2. Underexposure of the duplicate negative. 

In general, the fidelity of tone reproduction obtained at the five 
Hollywood laboratories studied in this survey is of very high quality. 
In most instances it is not possible to determine, by visual inspection, 
which, of the two prints made by each laboratory, is the print of the 
duplicate negative and which is the print of the original negative. 


The authors wish to express their appreciation to the five laborato- 
ries where the tests reported here were made. Thanks are due Mr. 
Ralph Westfall, who aided in compiling the data. 


1 IVES, C. E., AND CRABTREE, J. I.: "Two New Films for Duplicating Work", 
/. Soc. Mot. Pict. Eng., 29, 3 (Sept. 1937), p. 317. 

2 FRAYNE, J. G. : "Measurement of Photographic Printing Density", /. Soc. 
Mot. Pict. Eng., 36, 6 (June 1941), p. 622. 

3 KUNZ, C. J., GOLDBERG, H. E., AND IVES, C. E.: "Improvement in Illumina- 
tion Efficiency of Motion Picture Printers", /. Soc. Mot. Pict. Eng., 42, 5 (May 
1944), p. 294. 

4 JONES, L. A. : "On the Theory of Tone Reproduction With a Graphic Method 
for the Solution of Problems", /. Soc. Mot. Pict. Eng., 16, 5 (May 1931), p. 568. 

6 WEISS, J. P.: "Sensitometric Control of the Duping Process", /. Soc. Mot. 
Pict. Eng., 47, 6 (Dec. 1946), p. 443. 


MR. R. M. CORBIN: There have been many people who felt that sensitometric 
data were not too useful in this kind of work but I think they were led to that 
feeling by the use of time scale or long time exposures. You have to take into 
account a great many things that may happen to the sensitometric measurements 
in order to have them mean something. Only part of these factors are brought out 
in this paper. 

MR. JONES : In the case of images, does the image spread by reason of a higher 
density, and is there any distortion of line drawings? 

MR. CORBIN: With line drawings, an image spread or shrinkage occurs in 
accordance with the density used; of course, you can get cancellation by proper 
selection of the negative and print densities in the same way that you do in vari- 
able-area sound recording. 


W. C. EDDY** 

Summary. The three basic laws of television, i. e., mobility, flexibility, andin- 
stantaneity, are best represented in the requirements for a satisfactory lighting system 
for the video medium. 

While the fundamentals of lighting techniques for the stage and for motion pictures 
remain unchanged, the presently accepted system of remote-control overhead lighting 
in television relegates to one operator both the control and complete operation of the 
entire system. By flying the equipment from a unique gridiron, it is possible to keep 
the studio floor clear for camera operation and stage settings, while from an overhead 
observation and control position the lighting control engineer can, at a flick of the 
finger, create and recreate at will the lighting effects required at a given moment. 

Television studio lighting has in the past few years moved from the brute-force 
flat-illumination problem imposed by low-sensitivity cameras to the more specific 
requirements of the extremely high-sensitivity image-orthicon and orthicon chains. 
The need for high-intensity mercury-vapor lamps and over-all lighting levels of better 
than 1500 foot-candles incident has given way to a superflexible system capable of 
creating the artistic effects now required in studio work. The Tele-Lite of Television 
Associates, Inc., now installed in three postwar television studios, is typical of this 
new departure in remote-controlled illumination designed and tested in television 
broadcast studios. 

The advent of television, with its demands for a new high in con- 
tinuity of production, made apparent the need for a lighting system 
that would match the fluid spontaneity of the video cameras. While 
it was true that the basic requirements of stage and motion picture 
lighting remained unchanged, and that the type of light was the same, 
special equipment and special techniques for handling this new de- 
parture in lighting had to be created. ;V 

From 1934 to 1937, television lighting had limited itself to the art 
of pouring on front light in heretofore unheard-of quantities in a vain 
attempt to overcome the insensitivity of the early iconoscope cameras. 
Little or no choice was given or asked as to the methods used. Light 
and more light was the watchword, for it was apparent that light was 
helping to bring the television screen out of the shadows and in- 
definition of the scanning-disk days. 

* Presented Apr. 22, 1947, at the SMPE Convention in Chicago. 
** Director of Television, WBKB, Chicago, 111. 


During this era of development, lighting equipment in all its stand- 
ardized forms, and much that was not standardized, was inter- 
mittently brought into play in an attempt to overcome low mosaic 
sensitivity. Sun spots, Hollophanes, Broads, Inkies, Hi-arcs, and 
Kleigels could be found on stages in every experimental studio. 
Overhead in the acting area, massive arrangements of home-designed 
fixed ceiling units insured that no dimension of the stage would long 
be without its flood of Searing light, and, coincidentally, heat. While 
the design of these ceiling units varied widely from one broadcast 
studio to the next, they did have several characteristics in common. 

1. They were fixed in position, demanding that the subject be 
placed under the lights rather than bringing the light to the object 
being televised. 

2. They were, in general, a noncharacteristic type of illumination 
which did not and could not produce any but the faintest differential 
between high light and shadow, a major concession on the part of the 
lighting engineer to the erratic black-and-white response of the tele- 
vision screen of those days. 

3. Since they were fixed in position, there was little possibility 
of any but the crudest attempts at back and side lighting. Whether 
or not the pioneer televiewer would than have recognized or appreci- 
ated such techniques in that era of low fidelity cannot be guessed. 
Certainly the engineers of the period were more than glad to consider 
lighting in only one category, "quantity", and to overlook as well as 
discourage the so-called artistic attempts of the production group in 
introducing high light and shadow into the tele vision -picture business. 

This, briefly, is a resume* of television lighting in the late thirties: 
Immobility in equipment; insensitivity in camera equipment; with 
every licensee experimenting with new methods and devices to provide 
lighting levels of from one thousand to two thousand foot-candles 

In 1937, the field tests of the Radio Corporation of American and 
the National Broadcasting Company in the New York Metropolitan 
area brought television before the general public for the first time. 
It was then that the subject of controlled lighting was added to tele- 
vision's already extensive list of problems. From fan mail, telephone 
calls, and personal visits, it was immediately evident that the public, 
the ultimate consumer, wanted more than a high-fidelity reproduction 
of a kitchen chair or a flat-lighted engineering test pattern as its pic- 
ture fare from this new medium. The public wanted, and so stated 

336 EDDY Vol 49, No. 4 

in no uncertain terms, a reasonable facsimile of the lighting and 
camera work to which it had been educated by motion pictures. 
With the sensitivity of the cameras increasing by the week, with new 
techniques being devised in stagecraft, new technical equipment 
available in control rooms, it was easily apparent that if lighting were 
to stay abreast of the rapidly developing video field, a new and radical 
approach to the illumination problem had to be attempted. 

Television lighting, by popular acclaim oY popular criticism, be- 
came a point of issue. The wise broadcaster recognized the need 
and took steps to find an answer. With every competitor using 

FIG. 1. 

camera equipment of duplicate sensitivity and fidelity, it was patent 
that to the broadcaster who could couple the best in technical facili- 
ties with an artistic, well-composed, and well-lighted picture would 
go the best of competitive television contracts. 

For that reason a survey was made not only of the available systems 
but with a view to designing an ideal system as well. This analysis 
of an acceptable television lighting system brought out a list of ten 
basis specifications for which an answer had to be found. These 
requirements were: 

1. All lighting should be fully controllable from a remote position 
in the studio, preferably a catwalk above the studio floor. 


2. The basic lighting system should be made up of incandescent 
units because of the generally high camera sensitivity in this color 

3. All lighting should be controlled from one position and by one 
operator who could be trained and held responsible for both the artistic 
aspects as well as the engineering phases of the assignment. 

4. The accepted system should be capable of developing foot- 
candle readings of from 50 foot-candles incident to better than 1500 
to cover the wide range in sensitivity between the image-orthicon 
camera and its antithesis, the iconoscope. 

5. Sufficient light should be available in the system selected to 
light not only the stage in work, but to permit the presetting of at 
least one successive full-sized stage with normal illumination. 

6. A wide flexibility in type and character of light should be pro- 
vided to accommodate the characteristics of the several types of 
cameras being used or contemplated for television. 

7. This lighting system should be effective in any part of the studio 
with controlled back light, side light, overhead, and front light avail- 
able from any possible location on the set, throughout the course of 
the telecast. 

8. The unit should be light in weight, easily adaptable to new 
equipment, reasonable in unit cost, and simple in operation. 

9. Provision should be made to reduce light levels without re- 
sorting to dimmers during the long rehearsal periods required for a 
telecast. At the same time the angle of distribution, the relationship 
of high light to shadow, and the general effects to be achieved under 
broadcast conditions should remain unaltered to permit technical 
checking during rehearsal periods. 

10. The system should be controlled from an electrical switching 
panel that would permit instant as well as silent energizing of any 
or all units in the studio, without disturbing the camera circuits. In 
addition, a storage system of electrical controls should be incorporated 
in this switching panel to allow an accurate reproduction of the light- 
ing system and the position of the units devised during rehearsals 
to be repeated some hours later on the evening's telecast. In addi- 
tion to all this, the multiple-switch panel handling a possible peak 
load of 250 kilowatts should be absolutely silent and dependable in 

It was in answer to these specifications derived from actual studio 
experimentation that the normal stage-lighting equipment, common 



Vol 49, No. 4 

to motion picture studios, was in the main rejected. In the first place, 
the technique of lighting and the equipment used in motion pictures 
related to a picture system that could record and support high-con- 
trast scenes. In addition, individually operated units, by reason of 

their bulk and floor-mounted 
position, could not be expected 
to produce satisfactory lighting 
over an uninterrupted period of 
telecasting, without the benefit 
of frequent resetting and check- 
ing common to the motion pic- 
ture set. 

Third, there was not room on 
studio floors for the technicians 
and apparatus required for this 
normal type of floor-lighting 
equipment, without restricting 
the movement of cameras and 
microphone booms. Under test, 
it was proved that the indi- 
vidual ad lib lighting, as pro- 
duced by a series of independent 
floor operators, was at its best 
. problematical for commercial 
television, where "retakes" are 
impossible. If everything went 
right ; if each operator could be 
counted on to have his light, of 
the proper value, focused on the 
proper place, all went well- 
but let the slightest misplace- 
ment of the individual units 
occur, and the monitor screens 
of the shading console went into 
violent oscillations because of 
Jfe hot spots, or the picture resolved 

ilk itself into a panorama of dull 

gray and faded black. 

Based on actual television 
F IG . 2. studio experience, during both 


the early experimental period as well as the postwar commercial era, 
a remote-controlled system of television lighting was gradually 
evolved to answer the problems of television studio work. This 
system, controllable from a remote position through its arrangement 
of fair-leads, Fig. 1, and anchor blocks, Fig. 2, with the control center 
located in any convenient part of the studio, appeared to be a reason- 
able answer. 

FIG. 3. 

Second, by reason of the multiplicity of sockets (12 to a unit, 6 to a 
circuit), Fig. 3, lamps of any type or color coeffiicient could be used. 
In addition, a special adaptor made possible the installation and use of 
remotely controlled spots, Fig. 4, as well as special-effect lights with 
no further complication or restriction than that found in floodlights. 

In television studios, lighting problems were centered in one man. 
The lighting engineer was given not only the necessary control equip* 
ment, but the accompanying responsibility as well. Let us see how 
this system approximates the ideal. 

1. By relarrvping, this system can be operated in any range from 
50 to 2000 foot-candles per set. 

2. By proper arrangement of sufficient units on the gridiron, mul- 
tiple sets can be lighted simultaneously with back, side, and front 



Vol 49, No. 4 

FIG. 4. 

light, completely flexible and 
maneuverable in the hands of 
the control operator. 

3. The equipment is light 
in weight 35 pounds for a 6- 
kilowatt unit, which is less than 
6 pounds dead weight per kilo- 

4. It is cheap on a mass- 
production basis, and is fool- 
proof in operation. There are 
no motors to cause interference, 
no cooling system to leak, no 
complex controls to become 

5. A two-circuit system has 
been incorporated to permit 
half-light setups without vary- 
ing the color temperature of 
the lamps by resistance reduc- 
tion. By this means the studio- 
lighting level can be reduced 
for rehearsal periods at the 
same time maintaining an ac- 
curate reproduction of angle of 
distribution and high light to 
shadow relationship. 

In completing such a pack- 
age system of television light- 
ing, it was necessary that a 
master switching panel be de- 
signed to control the electrical 
circuits. After much experi- 
mentation with silent breakers, 
mercury switches, and other 
panel devices, a system of 
"momentary contact control- 
lers" was adopted, each switch 
controlling a remotely located 
breaker, two to each lighting 


unit. While it may seem that such a complex arrangement was an . 
unnecessarily expensive method of obtaining noiseless switching, it 
did have advantages which made its installation worth while. By 
using these magnetically held breakers, the switching system could be 
limited to low voltages, low-current circuits, thus reducing electrical 
"womps" on sudden changes of lighting load. 

At the same time all heavy-duty wiring to the lights came direct 
from the breaker panel to the unit, resulting in a large saving in 
heavy wiring costs. By adopting such a system, it was further pos- 
sible to incorporate a series of mechanical holding breakers which, 
when actuated, would allow operation of selected groups of units 
with the facility of a single unit. To complete this control board, 
a monitor receiver was installed as part of the assembly, to 
permit the operator first-hand analysis of the illumination on the 
stage in work. 

While it is evident that no system of lighting or combination of 
systems can, in themselves, insure satisfactory results, it is believed 
that the Tele-Lite system described has for the first time provided the 
telelighting engineer with equipment necessary to his assignment. 

With the increased contrast now available in modern television 
receivers, and the high sensitivity of today's cameras, it is to be ex- 
pected that new techniques in lighting, approximating the standards 
of motion pictures, will soon be displayed on receiver screens. By pro- 
viding a lighting system tailored to the requirements of the video 
arts, studio lighting under the guidance of a qualified technician is 
now able to parallel the technical advances of the engineering and 
production phases of television. 



Summary Lead-sulfide photoconductive cells developed during the war at 
Northwestern University show considerable promise in sound reproduction. These 
cells, in contrast with cesium- oxide phototubes used in present systems, exhibit a 
much higher signal-to-noise output and a lower impedance. The cell noise is not 
increased in the presence of background radiation. The frequency response is excel- 
lent and the sensitive surface is undamaged by high-light levels. As a result of the 
high infrared sensitivity of these cells, an indirectly heated exciter lamp has been de- 
veloped which operates with an ordinary 60-cycle filament transformer. Radio- 
frequency or direct-current heating of exciter lamps is thus not required. 

The detector used with optical sound tracks has been almost ex- 
clusively the cesium-oxy gen-silver photoemissive cell. The selen- 
ium-photovoltaic cell, such as is used in the modern exposure meter, 
has had limited use, particularly in Europe, but the high capacitance 
and consequent poor frequency response has prevented its use in 
high-fidelity sound systems. Recently, the so-called blue-sensitive 
photoemissive tube made with a cathode of cesium-antimony alloy 
has been used as a detector by several investigators. Unfortunately 
the extremely high sensitivity of this tube in the blue and ultraviolet 
is almost exactly offset by the feeble output of these radiations from 
tungsten-filament exciter lamps. The development of an exciter 
lamp with a high output in the blue region of the spectrum would 
make the performance of the cesium-antimony phototube much 
more impressive. 

Two new photoconductive cells made of thallous sulfide and lead 
sulfide have been recently released by the Government,. 1 These 
cells were developed during the war at Northwestern University 
largely under contract with the Office of Scientific Research and 
Development. The lead-sulfide cell is particularly adapted for use in 
sound reproduction by virtue of its high sensitivity, low noise, low 
impedance, excellent frequency response, and general sturdiness 
in the presence of background radiation. 

* Presented Apr. 21, 1947, at the SMPE Convention in Chicago. 
** Northwestern University, Evanston, 111. 



FIG. 1. Lead-sulfide cell. 


A detail drawing of one type of lead- 
sulfide cell is shown in Fig. 1. The lead 
sulfide is located on the inner wall of the 
envelope between the parallel conducting 
strips. The area of the sensitive surface 
depends on the application. For most 
sound systems areas of J / 4 X l /4 inch to 
*/2 X */2 inch are satisfactory. The resist- 
ance (in the dark) of the cells may be 
varied from about 0.1 to 10 megohms 
depending on the geometry and method 
of construction. Since no internal struc- 
ture exists inside the tube, internal micro- 
phonics are almost nonexistent. 

Spectral Response. The spectral re- 
sponse of the lead-sulfide cell is shown 
in Fig. 2. For comparison purposes 
the spectral characteristic of the cesium- 
oxide-silver phototube is shown. Ordi- 
nates are in arbitrary units. One of 
the most outstanding characteristics of 
the lead-sulfide cell is its high infrared 
response. It will be observed that in 
comparison to the threshold position at 
1.2 microns for the cesium-oxide-silver 
phototube, the lead-sulfide cell responds 
over a range of three octaves farther 
down in the frequency spectrum to 3.6 
microns. The response from 2.5 to 
3.6 microns is decreased somewhat by 
the absorption of the glass envelope. 

Frequency Response. A typical fre- 
quency-response curve from 30 to 10,000 
cycles is shown in Fig. 3. The modu- 
lated flux used for these data was 1 
microlumen from a tungsten filament 
at 2870 degrees Kelvin. The polariza- 
tion voltage applied in series with 



Vol 49, No. 4 

the cell and equal load resistor was 45 volts. The decrease in 
response at 10,000 cycles which in this instance is about 7 decibels 
is caused partly by the capacitance of the cell, base, socket, and 
connecting leads. 



60 . 



Response per unit radiant 
power in arbitrary units 


ead-sulfide cell with 
nonex glass envelope 

0.5 1.0 1.5 2.0 2.5 3.0 

FIG. 2. Spectral response of lead-sulfide cell. 


Signal In db 


Flux - I microlumen 
db= I microvolt 

Frequency in cps 
1 L 1 J 1 UjJ 

100 1000 

FIG. 3. Frequency response of lead-sulfide cell. 


Under identical test conditions cesium-oxide-silver phototubes 
deliver a signal 15 to 30 decibels lower over this range of frequencies. 

Cell Noise. The noise generated consists of two parts, namely, 
thermal or Johnson noise and current noise. For the polarizing 
voltages normally used in sound reproduction (45 to 90 volts) the 
total cell noise is not more than a few decibels above 1 microvolt. 
Actually the noise generated is a function of the area of the sensitive 

Oct. 1947 



surface and varies inversely 'with the square root of the sensithe 
area. The signal-to-noise ratio for a constant flux varies with area 
in the same way. Therefore the cell area should be no larger than 
is required by the optical system. The current part of the noise 
is also frequency dependent and decreases with increasing frequency . 

Response Versus Illumination; Background Effects. The re- 
sponse varies linearly with illumination up to values of 20 to 40 
foot-candles and thereafter seems to follow a square-root relation.' 

to grid 

Signal in db 

O.I 02 0.4 060.81.0 2.0 40 60&OIO 

FIG. 4. Signal versus load for lead-sulfide cell. 

When background illuminations are present from sources rich in 
the infrared, the signal output is decreased somewhat. In contrast 
to photoemissive cells in which the noise increases with background 
illumination the noise from the lead-sulfide cell is lowered. 

Signal Versus Load Resistance. The variation of signal output 
with load resistance for a constant input flux is shown in Fig. 4. 
Although optimum signal is obtained with a load resistance equal to 
cell resistance, the former may be varied considerably without a 
large decrease in signal. This characteristic may be utilized in 
sound systems without preamplifiers located near the cell to im- 
prove the over-all frequency response. By using a low-impedance 
input the effect of capacitance shunting at the higher frequencies 
may be reduced and a relatively long connecting cable to the 
amplifier may be used. 



Vol 49, No. 4 


The high infrared response of the lead-sulfide cell enables it to re- 
spond to sources of radiation at much lower temperatures than was 
possible with any previous photoelectric cell. In view of this char- 
acteristic an indirectly heated excited lamp has been developed in 
which the heating current is supplied by an ordinary filament trans- 
former. The details of one type of lamp are shown in Fig. 5. A lead- 
sulfide cell operating in conjection with this lamp whose filament is 
maintained at around 1500 degrees centigrade delivers a signal volt- 
age equal to or higher than that from conventional systems with an 

exciter lamp filament tempera- 
ture of around 2700 degrees 
centigrade. No 120-cycle hum 
is observed from the indirectly 
heated lamp. Conventional 
radio-frequency or direct-cur- 
rent heating of exciter lamps 
may thus be eliminated by using 

Metal sleeve 
Tungsten coil 

Insulating cement 

Filament (exaggerated) 

FIG. 5. Indirectly heated exciter lamp. 

the lead-sulfide cell as detector 
and the indirectly heated lamp. 


1 CASHMAN, R. J.: /. Opt. Soc. Amer., 36, 356A (1946). 

2 National Electronics Conference, Chicago, 111., 2 (Oct. 1946), p. 171. 


MR. LEWIS: As I understand it, this cell places before the film distributors or 
printers the question of whether to put out dye tracks or silver-salt sound tracks 
on their films. Is that correct? 

DR. R. J. CASHMAN: I don't believe I am in a -position to give a satisfactory 
answer to that question. The dye sound track has not been tried with this cell. 
It is well known that the dye sound track has a high transmission, around 8 /io 
or 9 /io of a micron. I have not found any data in the literature showing absorption 
or transmission characteristics in the region where this cell would come into its 
own. If there were an absorption band at 2 microns or 1 1 / 2 microns, then the con- 
trast might be great enough to make the cell work very well with a dye sound 

DR. J. G. FRAYNE: I should like to ask Dr. Cashman to tell us how the output 
versus the input compares with the standard setup. 

DR. CASHMAN: Thank you for reminding me of that. The output or signal 
voltage developed by the cell is a function of the intensity of the impinging jadia- 
tion. It is linear up to about 30 foot-candles. The relation from there on to higher 


intensities, like that of the sun, follows a square-root law, but up to about 30 foot- 
candles the cell is quite linear. 

There is another point in regard to background effects. No matter what the 
current is due to, the phototube increases its noise in proportion to the square 
root of the current through it. The current could be caused by background 
radiation. The lead-sulfide cell, on the other hand, decreases in noise with back- 
ground. The signal drops somewhat with background but the signal-to-noise 
ratio is reduced very little. 

DR. E. W. KELLOGG: From your description I didn't understand whether this 
is a thin film or not. You spoke of the area but I didn't understand the thickness. 

DR. CASHMAN: The thickness is such that the layer is about opaque. 

DR. KELLOGG: Very thin? 


DR. KELLOGG: I believe you said something about the name of the company 
that could supply it. 

DR. CASHMAN : I said that some of these types are made by the Electro- Voice 
Corporation in Chicago. Several other concerns are setting up manufacturing 
facilities to make the cell. 

MR. MORELOCK: Did I understand you to say that the exciter lamp is also 
commercially available? 

DR. CASHMAN : It is still being experimented with and it is not yet available. 

MR. GREEN: You made some comment with respect to the signal-to-noise 
ratio. Do I interpret it to mean that in running a variable-area track, which is ap- 
proximately half clear and half opaque, the lead-sulfide cell will be quieter? 

DR. CASHMAN : Yes, because the noise does not go up with the background 
it goes down. 

MR. W. S. MARTIN: What about the life of these cells in comparison with the 
old cells? 

DR. CASHMAN:' It is hard to say. We haven't had them long enough. Some 
have been in use since 1944. They are still just as good as when they were made. 

MR. MARTIN: Can the sensitivity of this cell be controlled very accurately? 

DR. CASHMAN : You mean in manufacture? 


DR. CASHMAN : The cell characteristics can be controlled quite accurately. 

MR. MARTIN : Have you done any experimenting with inserting a reflector to 
reflect the light after it passes through the film into the photocell instead of going 
direct from the exciter lamp after it passes through the film? 

DR. CASHMAN: I haven't, but I believe Mr. Van Niman has. 

MR. R. T. VAN NIMAN: We have used both lenses and mirrors. 

MR. E. I. SPONABLE: It might interest the membership to know that I spent 
about ten years on photoactive materials that change resistance on exposure to 
light. So far as I know, we at the Case Research Laboratory discovered the 
photoactivity of lead sulfide, antimony sulfide, and thallous sulfide. Back in 1918 
I remember that we were able to detect a man smoking a cigar a mile away. We 
thought that quite an achievement at the time. We also talked over a light beam 
a distance of some eight or ten miles. I am glad to hear that this interesting ma- 
terial has been rediscovered. 

348 CAMRAS Vol 49, No. 4 

DR. CASHMAN: If. I recall correctly, in your experiments, you used 60-inch 
mirrors or something like that. 

I was' instructed by the program committee to avoid historical surveys of this 
subject in order to save time. The pioneer work of the Case Research Laboratory, 
with which Mr. Sponable was connected, deserves the highest praise for its ex- 
ploratory work in this field. The laboratory is best known for its development of 
photosensitive thallous sulfide. The cells made with their material were not stable, 
however. This defect has been overcome in the modern thallous-sulfide cells. 
The Case laboratory also observed photosensitivity in natural lead sulfide (Galena) 
but this observation had been reported several times previously by other in- 
vestigators, for instance, Mercadier, U. S. Patent 420,884 (1890). The present 
lead-sulfide cell contains an activated synthetic preparation of lead sulfide. 



Summary. A magnetic track deposited between the sprocket holes and the edge 
of 8-mmfilm gives good-quality sound which can be added to any ordinary 8-mm film. 
Modifications of standard projectors for using this system are described. Perform- 
ances for speeds of 16, 18, and 24 frames per second are given. 

Although sound on 8-mm film has been considered in the past, re- 
sults were discouraging, and up to the present time no 8-mm sound 
projector has appeared on the market.t One difficulty with 8-mm 
films is the limited space available for a sound track. Fig. 1 shows 
the relative dimensions of standard films. The 35-mm film has a 
track about 100 mils wide; on 16-mm film it is about 80 mils wide. 
With 8-mm film the maximum track width is only about 30 mils. 
This track can be located at the film edge, either on the sprocketed 
side A or on the picture side B. 


More serious than the reduction in track width is the low 8-mm 
film speed. Table 1 compares the room available for storage of 

* Presented Apr. 25, 1947, at the SMPE Convention in Chicago. 
** Armour Research Foundation, Chicago 16, 111. 
t Apparatus using a separate disk phonograph is available. 



35-, 16-, and 8-mm sound. A 35-mm sound film running at 24 frames 
per second, goes through the soundhead at 18 inches per second, and 
the track is 100 mils wide. If we multiply the film speed by the track 
width we get a sound-storage index number of 1800. The 16-mm 
sound on the same basis has an index number Of 567, or roughly 31 
per cent as much. When we get to 8-mm film the index has dropped 
to 6 per cent for a 24-frame speed, and to only 4 per cent at a 16-frame 
speed. Experience shows that in going from 35-mm to 16-mm sound 

pi - 






1 6 MM 


FIG. 1. Relative sound-track dimensions. 

there is marked deterioration in quality. Considering that 16-mm 
sound with a rating of 31 per cent is not too far above the borderline 
for high-quality sound, the possibilities for 8-mm with a 4 per cent 
to 6 per cent rating seem discouraging. 

Type of Projection 
35-mm sound 
16-mm sound 
8-mm sound 
8-mm silent (1) 
8-mm silent (2) 


Sound Storage Index for Films 


per Second 


Index = 
Speed X 

Per Cent 







.*- - 7.2 





















Vol 49, No. 4 

If we choose to put the 8-mm sound track on the picture side, then 
we reduce the already limited picture area. Projectors would have 
to be modified for the smaller picture size. The possibility of adding 
tracks to old films would also be limited. We can avoid these dif- 
ficulties by locating the track on the sprocketed side. If we try 
optical sound there are photographic troubles which are indicated in 
Fig. 2. Frayne and Pagliarulo 1 have shown that film processing may 
cause uneven development of images that are as much as 30 mils from 
the sprocket holes. The uneven action of the developer at the edges 

FIG. 2. Region of uneven development. 

of the sprocket holes has been represented by vertical shading. There 
is also the possibility of action at the film edges, and this has been in- 
dicated by horizontal shading. Uneven development of this kind 
can cause sprocket-hole modulation even when the drive system is 

Economic problems also must be considered. Eight-millimeter 
photography has sacrificed quality in order to give the lowest pos- 
sible cost. If sound can be provided only by critical and expensive 
equipment and processes, the average amateur will not be able to 
afford it. 


Magnetic recording offers a fresh approach to the problem of sound 
for 8-mm projection. 2 Instead of an optical track of varying density 



or area, a layer of a newly developed magnetic material is bonded to 
the film in the space between the sprocket holes and film edge. This 
magnetic material has high coercive force and remanence, so that it 
may be magnetized in accordance with variations of magnetic flux in 
the 1 /2-mil gap of a recording head that rides against it. 

Fig. 3 shows one method for accomplishing this. With the selector 
switch in the record position shown, acoustic waves picked up by the 
microphone are amplified and fed into the magnetic head. Here they 
are changed into magnetic-flux variations which are recorded on the 
magnetic film track. The same head is used to translate the mag- 
netic record back into electrical energy. With the switch in playback 

FIG. 3. Magnetic sound-on-film recording system. 

position, these waves are amplified and fed into a loudspeaker. Al- 
though the record is "permanent" and will last for the life of the 
film, it may be erased quite readily by switching to the erase position 
and running a high-frequency alternating-current demagnetizing 
flux through the head. 

It is apparent that magnetic sound has a number of special advan- 
tages for the amateur : 

(1) Recordings can be made in the home, without special equip- 

(2) They can be played back immediately without processing. 

(3) Records may be erased and rerecorded. 

(4) Old films can be adapted for sound by adding a track. 

(5) Present silent equipment can be converted for sound. 

352 CAMRAS Vol 49, No. 4 


To demonstrate the possibilities of 8-mm systems, some conven- 
tional 8-mm projectors were converted for sound. Fig. 4 is a photo- 

FIG. 4. Eight-millimeter magnetic sound projector. 







: . ; 



^ ; 


]^ N \! 

1 : 


i ; 

\ \ 



1618 24i 







>0 100 1000 

FIG. 5. Over-all response of 8-mm system. 

graph of the equipment that you now hear. A flywheel and damper 
arms were added to the silent projector, This gives essentially the 



same mechanical system used by the company on its 16-mm sound 
equipment. It should be noted that if a 16-mm flywheel system is 
used on 8-mm, the energy storage is reduced to only 25 per cent, since 
the energy varies as the square of the velocity. A corresponding in- 
crease in flutter and "wow" should be expected, unless it is corrected 
by improved mechanical design. 

FIG. 6. Converted 8-mm silent projector. 

Of the projection speeds there are three possibilities to choose from : 
24, 18, or 16 frames per second. Eight-millimeter sound films which 
are made from 35- or 16-mm originals will most conveniently use 
24-frame projection. Best fidelity is offered by this speed. On 
the other hand, old silent films which have a track added should be 
run at their original 16-frame speed. It has been found by experience 



Vol 49, No. 4 

that most amateurs project their silent films at about 18 frames 
per second, since this "livens up" the action. Data for the 18- 
frame speed accordingly have been taken. On new productions in- 
tended for magnetic sound the amateur will have his choice of the 
higher fidelity 24-frame speed, or the more economical 18-frame 
speed (provided his camera can be set at either one). Frequency- 
response curves for the various speeds are given in Fig. 5. While 
not "high fidelity" the response compares with that of super- 
heterodyne radio receivers. Listeners have commented that both 
speech and music are excellent. It is interesting to note that in tests 




SCREWS <\. * 












FIG. 7. Adapter for 8-mm projectors. 

we have recorded as high as 10 kilocycles with a 2Y2-rnch-per-second 
film speed. This is not typical, of course, but it does indicate that 
there is room for future improvement. 

The present converted unit is not operating at its best because it 
uses an unregulated series motor. The addition of a governor, or 
the use of an alternating-current motor would decrease "wow" con- 
siderably. Another standard projector which has been converted is 
shown in Fig. 6. 


An attempt was made to design the simplest possible 8-mm sound 
adapter unit which could be applied to 8-mm projectors. One of the 
designs evolved is shown in Fig. 7. All of the essential parts are 



fastened to a plate which may be mounted on the projector with a 
pair of screws. The film comes down from the optical gate, and 
loops up past a pair of posts which take out most of the intermittent 
flutter. It then is pulled through a pair of friction shoes into the 
sound gate. The magnetic head is mounted in a recess in the sta- 
tionary shoe. Its high impedance of 9000 ohms at 1000 cycles allows 

FIG. 8. Adapter plate mounted on 8-mm projector. 

it to operate directly into the amplifier grid. A pair of posts between 
the drive sprocket and the friction shoes bend the film to provide 
compliance. All posts and shoes are grooved so they cannot scratch 
the picture portion of the film. The film compliance and friction 
blocks form a simple resistance-capacitance-type filter indicated by 
the electrical-filter circuit at the right. 


The adapter unit, mounted on a typical 8-mm projector, is shown in 
Fig. 8. The demonstration model operates under unfavorable con- 
ditions. It uses a series-type universal motor with poor regulation 
and V-belt coupling. Sprockets are driven by gears, and are of 
small diameter (12-tooth). The intermittent mechanism gives a 
fluctuating load on the poorly regulated motor. In spite of these 
faults (many of which could be corrected in a machine designed for 
sound adaptation) the projector does a creditable job for voice work, 
and gives quality that should be acceptable for such things as amateur 
titling and narrative. 


The author wishes to thank Ampro, Bell and Howell, and Univex 
for generously supplying equipment used in these tests. 


1 FRAYNE, J. G^, AND PAGHARULO, V. i "The Influence of Sprocket Holes on the 
Development of Adjacent Sound Track Areas", /. Soc. Mot. Pict. Eng., 28 
(March 1937), p. 235. 

2 CAMRAS, M.: "Recent Developments in Magnetic Recording for Motion 
Picture Film", /. Acous. Soc. Amer., 19 (March 1947), p. 322. 


MR. WILLIAM KRUSE: At what speed was the recorded talk played? 

MR. MARVIN CAMRAS: The voice was run at 24 frames and the music at 18 
frames, which is close to silent speed. 

MR. O. B. DEPUE: How do you apply the magnetic material on the film? 

MR. CAMRAS: The coating is in a liquid form and is flowed onto the film and is 
bonded into it. 

DR. E. W. KELLOGG: Have you done all of your applications of the coating 
material or is any commercial concern preparing that film? 

MR. CAMRAS: This has been done at the Armour Research laboratory. 

MR. KRUSE : What progress is being made toward commercializing the coating 
process? It seems that is the keynote to the usefulness of the whole thing. 

M-R. CAMRAS: You have to have both things simultaneously. You have to 
have projectors that will use the film and you have to have the film. The labora- 
tories that coat the film will want the market for it, and those who make the pro- 
jectors will want the film available. We hope that within a few months there 
will be some of this film available commercially and possibly some experimental 
equipment to use it. 



Summary This paper describes a method of re-recording 35-mm sound track 
to obtain a 16-mm track, which will reproduce synchronously with a 16-mm picture 
projected at 16 frames per second. The only necessary modification of a standard 
re-recording channel to perform this operation is the substitution of an 1800-revolution- 
per-minute synchronous motor for the standard 1200-revolution-per-minute motor on 
one 35-mm soundhead. 

Before the advent of sound recording on film and the establishment 
of 24 frames per second or 90 feet per minute as the standard of film 
speed in motion picture cameras and in sound-on-film recording 
machines, many millions of feet of 35-mm motion pictures were pho- 
tographed at the film. speed of 16 frames per second. Also, in the 
field of 16-mm amateur photography, practically all shooting has 
been and still is done at 16 frames per second. 

As is well known, in order to obtain a faithful pictorial representa- 
tion of the action photographed, the picture must be projected at 
the sa e speed as that at which the action was photographed. In 
other words, action photographed at 24 frames per second must be 
projected at 24 frames per second, and action photographed at 16 
frames per second must be projected at 16 frames per second. 

The problem has frequently occurred of adding synchronized 
music, sound effects, and commentary to these silent films shot at 16 
frames per second, and the question arises as to what is the most 
satisfactory and economical way of doing this without serious and ex- 
pensive modification of. the standard recording system designed to 
operate at a speed of 24 frames per second. 

First, let us consider the 35-mm films. Here we are faced with the 
condition that all 35-mm projection equipments, designed for sound 
reproduction, run at the standard projection speed of 24 frames per 
second. This leaves us, therefore, with only one obvious solution to 
the problem. In order to maintain the illusion of normal movement 
on the screen, it becomes necessary to double-print every other 
frame of picture and thus convert the 16-frame shooting speed into 
tjie equivalent of 24-frame shooting speed. 

* Presented Apr. 25, 1947, at the SMPE Convention in Chicago. 
** Warner Brothers Pictures, Inc., Burbank, Calif. 


358 HANSON Vol 49, No. 4 

The net result of this correction is normal speed of motion on the 
screen, but with an obvious lack of smoothness in the action. How- 
ever, the slight jerkiness that appears is much less objectionable than 
the ludicrous effect of all movement being one half faster than nor- 
mal, as is the case when a picture shot at 16 frames per second is pro- 
jected at 24 frames per second. After this pictorial correction is made, 
the recording of music, sound effects, and speech can then be added 
in the normal manner as is done with any standard 35-mm picture 
shot at the 24-frame-per-second speed. No changes in recording 
equipment or techniques are necessary. 

In the case of 16-mm films, however, the problem allows for other 
solutions because most sound-and-picture 16-mm projection machines 
are designed to run at either 16 or 24 frames per second. The problem 
could be solved by double-printing every other picture frame as in 
the 35-mm case, but since we have a 16-frame projection speed 
available, some scheme that avoids mutilation of the picture is much 
more desirable. One obvious way to do this would be to modify the 
sound-recording equipment to run at 16 frames per second. There 
are two serious objections to doing this. First, such a modification 
would temporarily put a standard recording channel out of commis- 
sion for normal recording and this, of course, would be prohibitively 
expensive in a commercial recording plant. Second, such a reduction 
in running speed would seriously affect the functioning of mechanical 
filtering devices designed to produce smooth film motion at the cor- 
rect 90-feet-per-minute operating speed of the recording and repro- 
ducing equipment. The following scheme is presented, therefore, as 
being the most satisfactory and. economical solution of producing 
sound-pn-film records which can be reproduced at 16 frames per sec- 
ond in a 16-mm projector. 

Let us assume that we are to add a combination of music, sound 
effects, and commentary to a 16-mm picture and that the final com- 
bined sound and picture print is to run at 16 frames per second. Also, 
assume that the commentary is to be recorded originally for the 16- 
mm picture and that the music and effects tracks are to be compiled 
from a stock 35-mm sound library. 

The procedure is as follows: The commentary is recorded on a 
standard 35-mm recorder, running at 90 feet per minute, while the 
commentator views the picture which is projected at 16 frames per 
second, using an interlock motor drive suitably geared to the pro- 
jector. The music and sound-effects tracks are compiled in the usual 


manner from footage measurements, allowing 3 3 /4 feet of 35-mm track 
for each foot of 1 6-mm picture. (Appendix A explains how this ratio is 
determined.) These tracks are then combined into a single 35-mm 
track by standard re-recording or dubbing methods. 

The resulting 35-mm dubbed track is then reproduced on a sound- 
head driven at 50 per cent above normal speed. This is accomplished 
by substituting an 1800-revolution-per-minute synchronous motor 
for the usual 1200-revolution-per-minute driver. By this operation, 
a frequency say of 6000 cycles in the recording is raised to 9000 cycles 
and the total playing time of the recording is two thirds of normal. 
This speeded-up sound track is then re-recorded on a 16-mm recording 
machine running at the standard 24-frame-per-second speed. During 
this re-recording operation, it is, of course, necessary to remove the 
usual 7000- or 8000-cycle low-pass filter from the re-recording channel, 
in order to pass the higher-frequency band being reproduced under this 
speeded-up condition. We now have a 16-mm version of our original 
recording which, if reproduced at the standard 24-frame speed, would 
be one-half octave too high in pitch and two thirds of the normal length 
in playing time. Now, if we reproduce this 16-mm sound track on a 
16-mm reproducer, running at 16 frames per second, the playing time 
and pitch are restored to normal and we have accomplished our ob- 
jective. As can be seen, the only modification that has been made 
to either the standard 35- or 16-mm recording equipment has been to 
equip one 35-mm soundhead with an 1800-revolution-per-minute 
driving motor, the remainder of the plant being left intact for stand- 
ard recording procedures. 

In order to study the effect of this type of re-recording operation on 
the frequency characteristic of the recorded material, the following 
test was performed. A 35-mm frequency test film having a constant 
percentage modulation for each frequency was first reproduced at the 
standard sound speed of 90 feet per minute through a re-recording 
channel having normal film-loss compensation. The substantially 
flat output from this film was fed to a 16-mm recorder running at 
sound speed and a 16-mm re-recording of the frequency film was made. 
The output from the 16-mm film was then measured on a Bell and 
Ho well 16-mm sound projector with the tone controls adjusted for 
optimum frequency response. The characteristic obtained is shown 
in curve, (A) Fig. 1. 

The same 35-mm frequency film was then reproduced on a sound- 
head running at 50 per cent overspeed and the output was fed through 



Vol 49, No. 4 

the same re-recording channel to the 16-mm recorder running at 
standard speed. The 16-mm re-recording thus obtained was then 
reproduced on the Bell and Howell projector, with the same tone-con- 
trol settings, but running at the silent speed of 16 frames per second. 
The output of this film was measured and the characteristic obtained 
is shown in curve (B), Fig. 1. 

As is to be expected, the 50 per cent overspeed recording results in 
a slight increase in film loss at the higher frequencies, while virtually 
no change, takes place in the low-frequency response. Compensation 




30 40 50 60 TO 80 100 


FIG. 1. Relative response of 16-mm sound track re-recorded from 
35-mm sound track. (A) 35-mm sound track and 16-mm sound track 
both reproduced at standard sound speed (24) frames per second. (B) 
35-mm track played 50 per cent over speed and 16-mm track repro- 
duced at silent speed (16) frames per second. 

for the high-frequency loss (6 decibels at 6000 cycles) can be provided 
for easily in the re-recording channel, but it is felt that for most com- 
mercial applications this compensation could be disregarded since the 
amount of equalization required comes well within the tone-control 
range of the average 16-mm sound projector and could be applied 

Actual commercial recordings produced in this manner substanti- 
ate this point and have demonstrated the highly satisfactory results 
that can be obtained by using the method described herein for adding 
sound to 16-mm pictures photographed and reproduced 16 frames per 


second with a minimum of modification to a standard recording and 
re-recording plant. 


One foot of 16-m'm picture has 40 frames. If this picture is pro-, 
jected at 16 frames per second, the length of film passing through 
the projector per second is 16/40 = 2 /$ foot. The standard projec- 
tion speed of 35-mm film is 90 feet per minute or l l / 2 feet per second. 
Therefore, the ratio of the length of 35-mm film projected at 90 feet 
per minute, to the length of 16-mm film projected at 16 frames per 
second, for equal projection times, is 1.5/0.4 or 3.75. 



When deciding on the width in direction of film travel of the illumi- 
nated area on the sound track ("scanning width"), a compromise 
has to be found between two contradictory requirements: more 
scanning width allows the passage of more light and gives a propor- 
tionally higher signal level for low frequencies, but it will cause an 
increased fall off at higher frequencies, where the scanning width 
covers an appreciable portion of one wave. Accordingly there exists 
for each frequency a scanning width with which the signal output is 

If the photoelectric amplifier is designed to provide sufficient am- 
plification for the highest frequency recorded, then the amplification 
for lower frequencies will be more than sufficient. Flat response can 
then be' obtained by attenuating these lower frequencies. Conse- 
quently, when deciding on the scanning width, we must consider 
only the highest required frequency. 

If we designate the reproducing scanning width by r, the frequency 
by /, the film travel speed by v, then the output signal will be pro- 
portional to 1 

sin r. 

^ (1) 

* Consulting Engineer, 1525 Teakwood Ave., Cincinnati 24, Ohio. 

362 FROMMER Vol 49, No. 4 

But the signal output is also proportional to the light flux in the 
scanned area, which in turn is proportional to the scanning width, so 
that the signal level at the frequency / will be proportional to 

*/ */ 

sin r sin r 

5f f 


To find the maximum of this expression, form the derivative 

, sin r 
-?- -"?' 


This derivative will equal zero where 

or r = 0.5 v/f. (5) 

Maximum output for the frequency / will be obtained at this 
scanning width. 

Numerical example : for a film travel speed of 36 feet per minute 
(7.2 inches per second) and for reproduction desired up to 6000 cycles 
per second, the scanning width optimum for highest signal input is 
0.5 v/f = 0.5 X 7.2/6000 = 0.0006 inch. 

A high photoelectric signal level is desirable mainly because it re- 
quires less amplification and accordingly that part of the background 
noise at the amplifier input, which is independent of the photoelectric 
signal, will be less amplified, or in other words, the signal-to-noise 
ratio will improve. That part of the background noise however, 
which is caused by thermal agitation in the photoelectric tube, is 
not independent of the photoelectric signal level, inasmuch as with 
increasing scanning width the average photoelectric current will in- 
crease and the thermal noise of the photoelectric tube will increase 
with the square root of this increased current. 2 Therefore in ap- 
plications in which the main concern is to improve the ratio between 
signal and the noise caused by thermal agitation in the photoelectric 
tube ("PE noise"), the optimum scanning width is the one at which 
the ratio between the photoelectric signal and the square root of the 
average photoelectric current is highest. 

The photoelectric signal is proportional to Eq (2), the average 
photoelectric current is proportional to the scanning width r, its 
square root to V r , consequently the ratio between photoelectric sig- 
nal and square root of average photoelectric current is proportional to 


sin r 

- ~ (6) 

V'r Sf r 4 

V V 

The first derivative of this expression is 

sin r f 

Or i / V // V , - / 7T/ /T\ 

r v cos r (7) 

which becomes zero where 

sin r 

Vif-V 1 = r- 1 /. cos 2 r 

Trf V 

or ~ r = 1/2 tan Sf r. 

This equation is satisfied by 

because tan 1.165 = 2.33 = 2 X 1.165, whence 

r = 1.165 


will give the highest P.E-signal-to-PJ3-noise ratio. 

Numerical example : For a film travel speed of 36 feet per minute 
(7.2 inches per second) and for reproduction desired up to 6000 
cycles per second, the optimum scanning width for highest P7i-signal- 
to-PE-noise ratio is 

0.370// = 0.37 X 7.2/6000 = 0.00044 inch. 


The actual background noise contains both PE noise and noise 
independent of the PE current. Accordingly the best scanning 
width is somewhere between 0.37 v/f and 0.5 v/f calculated for these 
two types of noise. 

The calculations in this paper were based on the assumption of 
constant light flux per illuminated area. In a further step this as- 
sumption has been dropped and the opening of the aperture stop has 
been taken as another variable. Publication of these more involved 
calculations however must be delayed till after publication of the 
underlying optical investigations. 


1 COOK, E. D.: "The Aperture Effect", /. Soc. Mot. Pict. Eng., 14, 6 (June 
1930), pp. 650-663. 

2 V. K. ZWORYKIN AND G. A. MORTON: "Television", John Wiley and Sons, 
New York, N. Y., 1940, p. 13, Eq 11.1. 



Summary A photoelectrically operated cuing system has been developed which 
avoids the difficulties associated -with lack of standardization on edge-notching of film, 
increases the ease of changing the cuing marks, and also makes feasible concurrent 
cuing of both light change and special effects. This system has been applied success- 
fully to a standard Bell and Howell Model J printer. However, the features dis- 
cussed are general and applicable to any type of film-moving system. Spots of white 
lacquer between the sprocket holes are illuminated by a small beam of light. The 
reflected light causes a discontinuity in the photoelectric-tube current which is amplified 
to operate a relay. The printer exposure channel actuates the normal light-change 
solenoid while the relay in the effect channel is used to initiate fade-in, fade-out cycles 
through the medium of a magnetic clutch which engages a rheostat in the lamp circuit. 
Practical operating experience and detailed mechanical features are described. 


The introduction of printer effects and the desirability of varying 
filter combinations during the printing operation have placed new and 
unusual demands upon the already inadequate method of printer 
cuing by edge-notching of the printer negative. The limitations, the 
lack of standardization, and the irrevocably permanent nature of the 
edge notch present very real problems. In some laboratories these 
problems are overcome by the use of a separate cuing film or tape. 
The preparation of this tape is often a costly and time-consuming 
process and its use is accompanied by added storage and filing prob- 
lems and introduces one more possibility for error in an already com- 
plex operation. 

Obviously, a dependable system of multiple-channel cuing which 
could be applied directly to the printer negative without damage to 
the film and which could be removed or altered with ease is greatly to 
be desired. The Moody Institute of Science has developed and is cur- 
rently using a system of photoelectric cuing for printing 16-mm 
films that gives promise of meeting these demands. A standard Bell 
and Howell Model / printer equipped with such a photoelectric cu- 
ing system is shown in Fig. 1. 

* Presented Apr. 22, 1947, at the SMPE Convention in Chicago. 
** Director, Moody Institute of Science, Los Angeles, Calif. 



In brief, this system utilizes for the cuing marks spots of white 
lacquer or other highly reflecting material approximating the size of 
a sprocket hole and applied midway between the sprocket holes of 
the printer negative as shown in Fig. 2. The pulse generated by a 
photoelectric tube as this spot passes a scanning aperture is amplified 


FIG. 1. Photoelectric cuing system installed on Model / Bell and Howell 


and used to operate a relay which initiates the desired function. By 
scanning both edges of the film, along the sprocket holes, two func- 
tions may be initiated in a simple manner. By applying somewhat 
more complex electronic selec- 
tion circuits it is possible to 
initiate more than one function 
on each channel by the use of 
single marks, double marks, and 
so on. The positioning and 
shaping of the cuing marks are 
not critical. Satisfactory results 
have been obtained with marks applied with a small brush. In any 
extensive use of such a system, however, the use of the more highly 
refined blooping techniques in applying the cuing marks would be 

FIG. 2. Placement of white cuing 
spots on 16-mm original film. 



Vol 49, No. 4 

The position of the scanning head on the printer is shown in the 
photographs of Figs. 1 and 3. In the latter figure the gate is open 
revealing the two scanning apertures between the film-bearing sur- 
faces. Fig. 4 shows the scanning head with the film threaded ready 
for use. 

FIG. 3. Scanning head with cover open. 

The drawing in Fig. 5 shows the details of the photoelectric scan- 
ning head. The light from a small flashlight bulb of the type having 
the lens molded into the envelope illuminates the film and the re- 
flected light falls upon a photoelectric cell. As a spot passes the 
scanning aperture, a short pulse of photoelectric current results due 
to the greater reflectivity of the white spot. The geometrical 

Oct. 1947 



arrangement is such that the light falling upon the photoelectric tube 
is that caused by diffuse, rather than specular, reflection, resulting in 
a more favorable signal-to-noise ratio. The signal-to-noise ratio of 
the system is established at this point, the background noise being es- 
sentially that due to modulation of the light by the sprocket holes. 

FIG. 4. Scanning head threaded and cover in place. 

This sprocket-hole modulation is minimized by blackening the areas 
of the backing plate behind the sprocket holes. Wear on the film as it 
passes through the scanning head is minimized by careful polishing 
of the bearing surfaces, but whatever wear results takes place outside 
of the picture area. 


The pulse from the photoelectric-tube circuit is led to a conven- 
tional resistance-capacitance-coupled amplifier consisting of four 

368 MOON Vol 49, No. 4 

stages, utilizing three type 5/7 tubes with a type 6V6 output stage. 
This output stage is transformer-coupled to a 500-ohm line terminat- 
ing in a conventional alternating-current relay. Feedback loops 
around the first two and also the last two stages stabilize the operation 
of the amplifier. 

For a film speed of 60 feet per minute, a spot width of approximately 
0.1 inch, and a scanning aperture approximately the size of the spot, 
the photoelectric pulse has a fundamental frequency of the order of 
60 cycles per second. The wave shape at the output of the amplifier 
is a single cycle of almost true sine shape. The time constants of 
the amplifier are adjusted so that full amplification is obtained at this 
frequency. In order to reduce the chance of the system respond- 
ing to spurious signals, the amplifier response is made to fall off rapidly 
above and below 60 cycles per second. Reducing the high-frequency 


FIG. 5. Drawings of scanning head. 

response eliminated trouble from tripping by transient electromag- 
netic fields set up when the light-change solenoid was operated. 
Some instability was experienced until an electronically regulated 
power supply was used for the amplifiers. With this highly regulated 
supply, very dependable and reproducible operating characteristics 
were obtained. It completely eliminated spurious operations of the 
relay resulting from power-line surges. 


A photoelectric cuing system as described could quite obviously be 
used to initiate many different functions pertinent to the printing 
process. For example, one of the simplest applications is that of 
controlling the light-change device of a standard printer. A sche- 
matic diagram of this is shown in the upper portion of Fig. 6. In this 
case the contacts of the relay controlled by the signal are connected 

Oct. 1947 



in parallel with the contactor normally actuated by the edge notch 
on the film. Aside from the fact that the initiating signal comes 
from the photoelectric channel rather than the edge-notch contactor, 
the operation of the printer is entirely normal. As the signal dura- 
tion, and hence the relay-closing time, at the output of the amplifier is 
of the order of a sixtieth of a second, this would apply only about one 
cycle of alternating-current power to the light-change solenoid. For 
positive operation it was found necessary to extend this time some- 



1 F 









| . | 








| . 



" - MOTOR 


1 ! 
l_ I 1 



FIG. 6. Operational diagram of cuing system. 

what by the addition of a supplementary direct-current relay having 
a capacitor across its actuating coil. 


The introduction of printer effects, such as fades and dissolves is 
another application of the photoelectric cuing system. Although the 
printer fade in this laboratory is now being achieved by lamp-voltage 
variation (a method which leaves something to be desired) the device 
is equally adaptable to other methods than the one herewith described. 
A schematic diagram- of the effects channel is shown in the lower part 
of Fig. 6. 



Vol 49, No. 4 

In the effects channel, the heart of the control lies in a magnetic 
clutch. After the initiation of the operation, a mechanico-electrical 
interlock completes the cycle. To assure positive start and stop, and 
to make possible precise timing of the cycle regardless of starting and 
stopping inertia factors, the drive motor is operated continuously and 
is coupled to the controlled function by means of a magnetic clutch. 
Such a clutch, quite simple in design and construction, has been 
found to engage and release very positively and, for all practical pur- 
poses, instantaneously. 

FIG.' 7. Magnetic clutch chassis with motor-driven fader rheostat. 

As shown in Fig. 7, a motor is coupled to a printer lamp fader rheo- 
stat through the medium of this magnetic clutch. The constructional 
details of the clutch are shown in the drawing of Fig. 8. The opera- 
tion of this interlock system is as follows : the amplified photoelectric 
pulse actuates a relay which closes a local circuit energizing the mag- 
netic-clutch coil. The motor is thus coupled to the shaft rotating it, 
and a cam-operated switch holds this clutch circuit closed once it has 
been initiated by the photoelectric pulse. At the completion of the 
fade-in or fade-out cycle, represented by a half turn of this shaft, 
the cam-operated relay opens the clutch cirquit. The next pulse 
would then complete the cycle, bringing the rheostat back to itsj 

Oct. 1947 



original position. The cam-operated relay is necessary to keep the 
magnetic clutch closed beyond the brief duration of the original 
initiating photoelectric pulse. 

FIG. 8. Sketch of magnetic-clutch details. 

The design of the fader rheostat is complicated by many dynamic 
factors such as persistence of vision, thermal inertia of the lamp fila- 
ment, and change of printer lamp resistance with temperature. A 
satisfactorily smooth approximation to an ideal straight-line curve 
(in terms of the change of density with time as viewed on the pro- 
jected film) has been made, utilizing 16 resistance increments per 
fade cycle for Kodachrome duplication. 


Summary The relationship between the area, shape, and fitments of a sound 
studio or auditorium and their effect on sound quality is discussed. A probable 
evolution of music appreciation, from the pentatonic and whole-note scales of certain 
primitive peoples to harmony, as we know it today, is traced. Various methods of 
acoustical treatment of enclosed areas are described and illustrated. 

Sound is a form of pulsating or vibrating energy and in this dis- 
cussion we are concerned with its behavior in an enclosure or re- 
stricted space. The behavior of sound energy in space, be it in a 
theater, studio, living room or bathroom, is acoustics. The source of 
the sounds may be a vibrating string, membrane, column of air, or 
vocal chords. The air as a medium coupled to the source carries the 
vibration throughout the space until attenuated by distance, or 
absorbed by boundary surfaces. 

Sustained sounds thus will fill a space and the intensity will grow 
by the addition of the various reflections until that steady state is 
reached when the energy of the source is balanced by the boundary 
loss, or absorption. If the source is suddenly stopped the residual 
sounds continue to reflect and to die away because of absorption. 

The time taken for the sound level to drop 60 decibels is called the 
reverberation time. Some modern structures with hard surfaces 
have been known to have reverberation times in excess of 10 seconds. 
When we consider that we normally speak about three syllables per 
second and this extreme would mean thirty syllables running around 
the hall at the same time, we see what utter confusion and unintel- 
ligibility can result from excessive reverberation. Music in the same 
place would sound as though the loud pedal were on all of the time. 

"It is probable that the growth of modern music and our apprecia- 
tion of it have been molded by the architecture of our buildings. The 
music of most primitive civilizations developed out of doors, as the 
homes of these early cultures were generally in tropic or semitropic 
climes. Their music has survived in the Near East and in the Orient 

* Presented Feb. 19, 1947, Atlantic Coast Section, SMPE, in New York. 
** Johns-Man ville, New York, N. Y. 


largely in pentatonic and whole-note scales. Harmony as we know it 
cannot be obtained using these scales. They are best suited to heavy 
octave melodies, or staccato tunes for flute and strings built over the 
various drum rhythms. 

An example of this use of ancient scales is in Japanese music today. 
Their traditional music is played out of doors, in shrines open at the 
sides, or in houses constructed of thin wood, bamboo, and paper so 
flimsy and absorbent as to simulate outdoor conditions. However, 
as the Japanese have gradually adopted the fireproof construction of 
the west, their modern music has likewise begun to make use of our 
scales and to take on the style of western music. 

In the Christian era the march of civilization was toward the tem- 
perate zone of Europe. During the early centuries the church be- 
came the seat of culture and the main source of developing musical 
tastes. In the main, the churches, temples, and cathedrals were of 
much greater volume than other structures and they were enclosed 
against the more rigorous climate. Stone replaced wood as more 
enduring materials were used. 

These factors aided in prolonging reverberation of sounds because 
reverberation depends on the amount of space enclosed and on the 
hardness of the surfaces. The notes overlapped. The more primi- 
tive scales when used in such places provided dissonances which were 
unpleasant to the ear. The Gregorian chants arose. The diatonic 
scale came into popular use, particularly in early operas. The hurdy- 
gurdy grew into a harpsichord, then into the piano and organ. Bach, 
Beethoven, Brahms, Stravinsky, Debussy, Ravel; changing scales, 
new instruments, great composers and new architecture. 

There are many of us who remember the early silent motion pic- 
tures where the sound effects consisted mainly of a piano and the 
quality of the music was of little value except to cover up the sound 
of the projection machine and to fill the time of changing reels and 
repairing film breaks. The legitimate theaters, concert halls, and 
churches inherently either had good or bad acoustics and little was 
done about it. 

"Acoustics" was about as vague a term as politics, as little known 
and frequently as aimless. During this period miles of wire were 
strung up in auditoriums all over the country and some of it still 
exists. While it was one of the early attempts at acoustical treatment 
it is known that it did no good at all. 

In the early part of this century Professor Wallace Clement Sabine 

374 DUNBAR Vol 49, No. 4 

of Harvard University put the acoustics of buildings on a scientific 
footing by establishing the relationship between reverberation time, 
room volume, and sound absorption. 

The subsequent treatment of many rooms and the use of articu- 
lation tests have determined that the optimum value of reverberation 
for a room is a function of its volume. Also, it was learned that a 
room used almost entirely for music should be somewhat more lively 
than one of the same volume intended primarily for speech. 

As long as we depended mainly oh the power of our own voices and 
musical instruments, most of our auditoriums were fairly satisfactory. 
The bad ones were usable even though they were annoying. But 
when the audion valve, vacuum tube, and triode came into being the 
course of the entire history of sound changed. We could then am- 
plify the sounds we knew, thus putting many times the. sound energy 
into the same enclosed space. In this way the old acoustical defects 
were multiplied. 

Sound came to us in the form of radio-receiving sets, public-address 
systems, and sound motion pictures. People accepted the deficien- 
cies of these new forms of education and entertainment as long as the 
novelty remained. But soon they were complaining about the old 
halls that boomed and growled with reverberation and echoes. 'They 
wanted something better. And as the sound industry improved 
acoustics had to improve with it. 

We had had experiences in our homes with the sound-deadening 
effects obtained by the use of heavy carpets, draperies, and similar 
materials; however, little was known of the very few commercial 
acoustical materials existing at that time. So the attempts to 
deaden the noisy rooms and halls were generally made with home 

Accordingly, upholstered seats, carpets, and acres of draperies 
made their sudden appearance. Draperies ranged through theatrical 
gauze to monk's cloth and heavy velour folds alone, or applied over 
all kinds of felts, jute, and cattle hair. Quilts, blankets, and rugs, 
too, were hung up in attempts to subdue the bouncing sounds. 

The same unstudied treatment was given to early radio studios, 
recording studios, and sound stages. The pickup and recording 
equipment were just about as rough in those days as the rooms in 
which they were used, and the noisy scratch of early synchronized 
films would make poor entertainment today. 

All of this made a growing industry conscious of the need to learn 

Oct. 1947 



more of the nature and behavior of sounds ; speech and music in par- 
ticular, for these are the intelligible noises which give instruction and 

It is obvious that the cure for reverberation is sound absorption; 
however, this cure can be excessive in two ways. Materials may be 
used which absorb only a part of the sound spectrum, or an excess of 
material may produce too much over-all absorption. 

A room treated so as to produce selective absorption will give a 
very unnatural result. If the high frequencies are heavily absorbed 

Type of theater using treated domes, heavy ornamentation, and 
upholstered seats. 

the room will be boomy and the voice and instruments hardly dis- 
tinguishable because of the loss of overtones. The degree of absorp- 
tion should be about the same at all audible frequencies to maintain 

Many studios have been treated with material intended mainly for 
high-frequency typewriter or knife-and-fork noise, and the musicians 
have complained that these spaces are too dead, or too boomy. 
These studios have been corrected by taking out some of the high- 
frequency absorbent until a balance was secured. 

376 DUNBAR Vol 49, No. 4 

The second excess is too much over-all absorption, making the room 
approach zero-reverberation or "free-field" conditions. Here we go 
back to outdoor conditions for which our music is not suited. An 
acceptable balance must be secured based on study of old halls, long 
used and liked, or new ones thought to be better because of public 
acceptance of the results obtained in them. Knowing the upper and 
lower limits has been a great help in setting up proper reverberation 
times for various spaces. 

When a room is to be used for recorded sound, both the originating 
studio and the hall should be treated slightly more than the rule calls 
for because of the additive effect of their reverberations. This is 
true of sound stages, motion picture theaters, and recording and 
broadcast studios. 

Another fault commonly found in older theaters was echoes, or 
delayed repeats, of the original sound. These were generally caused 
by large domes, curves, and large flat surfaces. If a curved ceiling 
had focuses at the floor level or even multiples of this distance, dis- 
agreeable echoes were more than likely to spoil the seating down the 
center of the room. Most modern designers avoid these shapes and 
the domes of many of the old auditoriums have been heavily treated 
or new ceilings suspended under them with curves to kill the echo. 
An example of the use of a treated dome is shown in Fig. 1 . 

The modern theater is now generally constructed of splayed sur- 
faces which tend to straighten the path of reflected sounds passing to 
the rear. The back wall is not curved to focus near the stage unless 
it is very heavily treated, and upholstered seats are used to act as 
compensation absorption with small audiences. This last item keeps 
the reverberation from varying too much between full and empty 
conditions and does not require such a wide swing in gain to operate 
the horns. 

The recording studio is a special problem in acoustics. First 
the sound should be so well distributed as to require only one micro- 
phone, or a very few in any case. If possible the reverberation 
should be quite flat over the frequency range except for the high end 
where it should rise. 

The scratch and circuit noises are still intense enough to make very 
wide range recording a little unpleasing. If the high frequencies, 
that is, the overtones, timbre, and brilliance, can be built up exces- 
sively by reverberation and attenuated electrically back to normal 
level, a high signal-to-noise ratio will be maintained and the scratch 

Oct. 1947 ' SPACE ACOUSTICS 377 

considerably subdued. Such a treatment must be done carefully 
with as much attention to reflective surfaces as to the location and 
kind of absorbing surfaces. 

Reflective materials are, generally, the floor which must be hard 
enough for its traffic, plastered surfaces, plywood, and transite. 
Plaster, plywood, and transite may be formed into polycylindrical 
and faceted or prismatic surfaces. These scatter sounds for better 
mixing. The deeper the offsets, the lower the frequencies which are 
affected. The sound-absorbing materials generally have a base of 
rock wool which will be covered by membranes of thin plywood, or 
perforated transite, to avoid too much high-frequency absorption. 

FIG. 2. A Reeves Studio in New York City. 

For the same reason, great care should be taken to use sparingly such 
materials as carpeting and draperies. (Note the serrated panels 
in the ceiling of Reeves' studio in New York which is shown in Fig. 
2 and in the control room of this studio in Fig 3.) 

The problem is virtually the same for the frequency-modulation 
studio, but it is not quite so rigorous for the more general amplitude- 
modulation studio because of the millions of limited-range receiving 
sets already owned by the public. Most of these run almost entirely 
on bass notes and almost twenty-four hours a day. However, as 
equipment is improved, all of these amplitude-modulation studios 
will have to be adjusted for high-fidelity sound. 

There is another problem connected with all studios and places 
where we listen to sound. This is background noise. Practically 
all theaters are in urban areas with plenty of traffic noise all around. 

378 DUNBAR Vol*49, No. 4 

and their problem has been solved by omitting windows, using heavy 
walls, and deep vestibules, heavily carpeted, and lobbies with double 
doors. We must expect to endure some objectionable background noise 
when we sit in the middle of a thousand or so people. Fortunately, 
we become psychologically deaf to their presence when the show is 
interesting. This is not true of a real studio. A recording or broad- 
cast that carries with it traffic and fire-engine noise, or the sounds 

FIG. 3. Control room for studio shown in Fig. 2. 

from other spaces in the building such as presses, elevators, and 
plumbing is not interesting. The background noise in a studio should 
be such that the intended silent moments are really silent and not 

Such conditions can only be attained by sound isolation. This 
work bears no relation to the acoustics of the room itself and is a mat- 
ter of construction and location. A single studio may be built far 
enough in the country to avoid all noise except thunder and airplanes, 
yet it still must be isolated from its control room and recording room 
to prevent feedback and pickup of machine noise. Most studios are 

Oct. 1947 SPACE ACOUSTICS 379 

built in the cities to be accessible to talent, service, and other requi- 
sites. Here they must contend with other building noises and 
sometimes subways, heavy traffic, and near-by manufacturing. 

There is no such thing as absolute soundproof ness; at least not on 
earth. A body at absolute zero floating in an absolute vacuum 
might be said to be absolutely soundproofed. Under special con- 
ditions we are able to hear sounds at some pitches down to a point a 
little below zero decibels, which is a sound intensity of 10 to 16 watts 
per square centimeter. However, as a person's body makes a noise 
of about 14 decibels just doing its regular work of pumping, digesting, 
and breathing we do not often experience quiet below that. 

Ordinarily a studio is considered good if the background level can 
be held to the twenties of decibels. Naturally, more soundproofing 
will be needed in a noisy than in a quiet location. Suppose we build 
a wall which will attenuate sound 40 decibels. If the level around it 
is 70 decibels, the level inside the walls will be 30 decibels; if the 
level outside rises to 90 decibels when a truck goes by, the inside 
level goes up to 50 decibels. In such a location more soundproofing 
will be required. 

The foregoing indicates clearly that a proposed studio location 
should be examined before construction is contemplated in order to 
determine the problems to be expected from extraneous sounds. This 
survey should include observations of the background noise level, 
the amount of vibration in the structure, its mass and rigidity, and 
its allowable floor loading. 

A building on Third Avenue in New York was recently examined in 
order to determine if it could be used for recording studios. Because 
of the scarcity of space a thorough examination was made of it. It 
was an old building with lightweight wood joist floors and ceilings 
which could not stand the weight of studio construction. The 
ceilings were so low that if we had used the necessary space for iso- 
lation and treatment, there would hardly have been room left for a. 
bass viol. And besides, every time a Third Avenue elevated train 
passed by the whole building shook and the noise level rose to a point 
which made conversation impossible. We had to advise against 
signing the lease. 

The weight of the structure is very important as the heavier the 
walls and slab construction, the better it resists vibration and passage 
of sound. A 12-inch brick wall is. much more soundproof than a 
stud partition, as everyone knows. 



Vol 49, No. 4 

The majority of the better studios are constructed of the "room- 
within-a-room" principle. The interior surfaces of the studio are 
"floated" away from the actual building floor, ceiling slabs, and 
structural partitions. This "floating" is accomplished by means of 
felt or spring isolators which carry or support the inner surfaces of 
plaster and floor slab. An example is shown in Fig. 4. Here the 
walls and ceiling have already been isolated, sound-absorbing panels 
installed, and the floated floor is being constructed . 

r " , 
FIG. 4. Studio in process of being isolated. 

Entrances to the studio are accomplished through soundproof 
doors. The ventilating system is flexibly connected to the room and 
c e ducts are lined with sound-absorbing material. The electrical 
.conduits must be flexible, too, where they enter. Heavy double 
glass is used for observation windows. Each glass is of different 
thickness or tilted out of parallel with the other to reduce trans- 

After all of this is done the isolated studio is ready for acoustical 
correction of the interior which means the control and distribution of 
reflected sounds within it. This is accomplished, in the manner al- 
ready discussed, by sound-absorbing material for different pitches of 
sound and reflecting material at suitable angles for distribution of 

Oct. 1947 



sound. An example of a completed studio is shown in Fig. 5. This 
is Studio 8-H of the National Broadcasting Company in New York. 

The associated control room is soundproofed from the studio to 
prevent feedback. It need not be quite so soundproof against out- 
side noise as the studio itself, but it must have nearly perfect listening 
conditions for proper monitoring of a program, judging quality, mi- 
crophone placement, and poor performance. 

FIG. 5. Studio 8-H of the National Broadcasting Company in Radio City, 

New York. 

The higher the fidelity of the program, the more necessary it be- 
comes to try for acoustical perfection in both the studio and the con- 
trol room. A place used primarily for playing records is relatively 
unimportant compared to a studio originating frequency-modulation 

All of us are interested in television. A studio used for telecasting 
is in some ways quite different from those used for broadcasting and 
recording ; yet they are similar in many respects to the present sound 
stages for motion pictures. Here, again, the first requirement is 
isolation from extraneous noise to prevent transmission of something 
foreign to the program. 

382 DUNBAR Vol 49, No. 4 

In television as in sound motion pictures, two senses are being 
catered to, the sound and the vision must match. If an outdoor 
scene is being shot the sound should not be reverberent as though 
coming from a rain barrel; and conversely an indoor scene must 
carry a certain amount of room acoustics to make it sound real. The 
first condition is secured by covering practically all of the interior 
surfaces with a very efficient sound-absorbing material, to allow as 
little echo and reverberation as possible. In the second condition, 
when shots require reverberation, the reflection of sound is usually 
accomplished by reflective surfaces of sets and flies, or by adding 
reverberation through chambers or mechanical devices. 

The reverberation chamber is a room with very hard surfaces. 
The echo chamber is one where a desired length of sound path may be 
used between microphone and speaker. One mechanical device is com- 
posed of a series of long damped springs with the alternate ends mass- 
loaded and clamped for reflection of mechanical vibration induced 
by the signal. These devices may be used as a parallel signal path 
so as much of their effect may be added to the signal as is desired. 

A television studio should have a- ceiling height capable of ac- 
commodating lighting. Usually catwalks are used to place and 
mount special lights. The ventilating system must be adequate to 
handle the increased load caused by the heat from the intense lighting. 
The whole effort is to bring as many scenic conditions as possible 
under a protective roof and within soundproof walls and close to the 
elaborate equipment needed for pickup and telecasting a program. 

In all of these enclosed spaces we are attempting to produce pro- 
grams and entertainment which will be the most natural and pleasing 
to the greatest number of people. For this reason sound coming 
with television and with motion pictures, too, should match the scene 
conditions. If the actors are evidently in a living room their voices 
should sound as though they were in a normal-sized room and not as 
if they were in the bathroom, or out in the woods. Music, too, should 
have its normal reverberation and not give the impression that the 
violins are stuffed with cotton. 

This whole science of architectural, or space acoustics, as we have 
called it, is a rather simple thing. It is just a matter of common 
sense to have our studios treated to give them natural sound and to 
have our theaters designed so that everyone in them hears the sound 
as it should be. If the rules previously stated are followed there 
should be no reason why this goal cannot be attained. 

Oct. 1947 SPACE ACOUSTICS 383 


MR. RIDGEWAY: What can you tell about the latest methods of motion pic- 
ture studio sound treatment? 

MR. JAMES Y. DUNBAR: We do not get as much exposure to that problem on 
this coast as they do in the west, but the film sound stage involves the same proc- 
ess of treatment as the television stage. One of the important things is to keep 
extraneous noise out of it. It should be fixed so that extraneous noise is not added 
to the recorded sound. Of course, a great deal of sound is dubbed in at places 
suitable for sound recording after the picture is shot; but where it is not, the 
acoustical treatment has to be such that one can obtain a wide variety of scenes 
and sounds under one roof, from outdoor effects down to room effects. 

The treatment should be the most efficient absorbent obtainable to prevent un- 
natural effects. If you have a place that has only high-frequency absorption and 
is rather boomy, you know from the sound that you are not outdoors, even though 
the scene tells you that you should be. 

MR. RIDGEWAY: Have there been any recent developments in materials used 
in motion pictures, other than rock wool ? 

MR. DUNBAR: The use of rock wool continues, although its fabrication has 
changed a bit. One way of absorbing the lower frequencies is to make the treat- 
ment thicker and to space it out from the hard walls or ceiling. This has been 
taken to quite a length in what they call anechoic or nonechoing chambers. I 
think the early work on these was done by Meyers in Germany. This process 
consists of cones or pyramids of rock wool three or four feet in length. The entire 
room surface is covered. It is an excellent treatment, but the cost is more than 
could be allowed in the construction of sound stages. It is only in special testing 
chambers that such a treatment is used. It was used in one of the test rooms in 
the Brooklyn Navy Yard. 

MR. BARNES: Is there any preferred ratio of length or breadth to height, in 
studio construction? 

MR. DUNBAR: There has been a great deal published on that, and I say, "No", 
within certain limits. I have known some people with excellent space for a studio 
who wanted to restrict it to a ratio of 2:3:5 or 2:3:6, or 2:3 '4. There is no partic- 
ular reason for this. 

Through the long period in which we have acquired our sense of music apprecia- 
tion and sound, we have become accustomed to living indoors and most of our 
experience has been indoor experience in rooms of varying sizes. We have gotten 
away from the Greek amphitheater and expect to notice some room effect. 

In a room of a size which is peculiar, like a very large one with an excessively 
low ceiling, the poor room proportions would be apparent to a blind man. A 
blind man can get some sense of the dimensions of a room by walking into it and 
hearing reflected sounds. We may not be so conscious of this as a blind man, but 
we know a room has to be proportioned somewhat near the sizes we are used to in 
our living. If studio size is within this range it is sensed as all right; if not, cus- 
tom and binaural hearing will probably tell us that it is peculiar. 

MR. BARNES: Would a room that had a ceiling height more than the width be 

MR. DUNBAR: It is likely to be unless treated as a room turned on its side, 
which it could be. And it would probably still be troublesome! 

384 DUNBAR Vol 49, No. 4 

MR. C. R. KEITH: Nowadays, 16-millimeter motion pictures are shown in 
classrooms and other rooms not particularly designed for reproduction of sound. 
I wonder if you have any particular suggestions that might apply to such 

MR. DUNBAR: More and more, I have had architects ask for mineral treatment 
of classrooms, and I think a great many of the modern schools will have treat- 
ment in the classrooms to adapt them to the use of visual aids and use of motion 
picture 16-millimeter films. Most of the newer schools are having acoustically 
treated classrooms specified. 

MR. KEITH: Is there any treatment of existing rooms which might be practical, 
or is it usually necessary to revamp completely the whole room? 

MR. DUNBAR: No, that is not necessary. Many materials can be glued, clipped, 
or nailed on. And it is possible to secure just as good a treatment in an old room 
as a new one. 

MR. LEWIN: In cases of small studios, used for a commentator talking into a 
microphone, is the best treatment to make the rest of the room as dead as possible? 

MR. DUNBAR: No, a muffled effect will result. It is better in a small room to 
use considerable low-frequency absorption. A male commentator has a low voice, 
and the common treatments have no effect on the fundamental. If you do not 
absorb an excessive amount of the highs, a richness, distinctness, and clarity to 
speech stand out. 

MR. LEWIN: Would it not appear that when a person is talking very close to 
the microphone practically all of the sound is being picked up directly? 

MR. DUNBAR: Practically all of it is. 

MR. LEWIN: What is picked up from the wall should be less important, and 
might as well be deadened. 

MR. DUNBAR: But if you have a perfectly dead room condition, talking into a 
microphone is not too pleasing to the speaker. 

MR. SAWYER: Would you care to comment at ail on tests that you might 
apply to an important room such as broadcast studio or recording studio? In 
other words, do you depend pretty largely on your calculations in advance, or 
when a job is done, do you make some measurements and modifications? 

MR. DUNBAR: We have made quite a few reverberation measurements before 
and after treatment, and after measurements are generally fairly close to the 
calculated. The treatment that is used has been laboratory- tested for the co- 
efficient, so we know how much absorption we are putting in. There may be other 
things like the calculations of, say, a plaster wall which looks rigid. It may be 
very thin and quite flexible and therefore very absorptive, and we may have mis- 
estimated it in our calculations. An actual test will show that up. 

I might ask Mr. Gurin how close he came on his last measurements. Would 
you mind telling us? 

MR. GURIN: Ten per cent. 

MR. DUNBAR: That is close. I wish we could always do as well. 

MR. SCHLANGER: Much has been said on the subject of no absorption treat- 
ment at all. In small theaters, to what extent is that possible today? 

MR. DUNBAR: I might point out that when La Salle Playel was designed in 
Paris, all sound was focused back to the audience ; but one item that was over- 
looked was reverberation. Treatment had to be installed afterwards to reduce 

Oct. 1947 SPACE ACOUSTICS 385 

the excessive reverberations. This treatment consisted largely of absorption 
applied on the back wall. 

MR. SCHLANGER: How about a theater of from 300 to 600 or 700 seats? 

MR. DUNBAR: I think if you hold down the unit volume and maintain some 
nonparallelity of the surfaces to straighten the reflecting paths a bit you would 
still have to have some back-wall treatment. If the volume is excessive in pro- 
portion to the number you are seating, a large amount of absorption must be used 
to hold the reverberation down. 

MR. SCHLANGER: How about not letting the paths come back to the source? 

MR. DUNBAR: This can be helped by sloping the back wall. If the back wall 
focuses toward the audience, the reflected sound meets with the oncoming sound. 
You might have sufficient lag to cause a disturbance, or echo in such an area. 

MR. SCHLANGER: Assume you could reflect the sound downward. 

MR. DUNBAR: If you have heavy absorption on the floor that would help. 

MR. SCHLANGER: The audience would help, too. 

MR. DUNBAR: Yes, it would. Another way would be to slope the back wall up 
and put heavy sound absorption on it; the second reflection is then negligible. 
Also you might scatter it by broken surfaces and absorption for good results. 

MR. LEWIN : In a sense, I think you have given us two concepts here. They are 
somewhat contradictory, because you outlined the way you treated a theater and 
the desire for having absorption on the back wall, on the ground that the sound 
presumably is headed in one direction and should not go back the other way, but 
when reference was made to recording studios, you gave us the inference that the 
idea of a live-end and dead-end was wrong, which one might assume to mean that 
you would want to use that studio one day with the orchestra on one end and 
another day with it on the other. 

Now, if the absorption on the back wall of a theater is good, why isn't a live- 
end-dead-end studio desirable on the ground that it is a large studio where you 
have audiences in one end and an orchestra in the other? 

MR. DUNBAR: To keep a theatei from being a live-end-dead-end, you put 
treatment on the back wall. The front wall is already dead, for you have con- 
siderable absorption on the screen side. To go back to the legitimate theater with 
flies and props and curtains, the proscenium area is very dead. But if your 
highly directional horns with considerable intensity are pointing toward an un- 
treated hard wall, you can expect trouble. You may expect a lot of trouble in a 
studio if you start shooting such a horn around and try picking it up on a micro- 
phone. In a motion picture theater you do not want to be conscious of the acous- 
tics of the theater itself but to hear the sound as though it were coming from the 
picture shown. The acoustics of a studio, however, are a part of the sound picked 
up for recording or broadcasting. 

MR. JORDAN : A question occurs to me as to the effect of acoustics. I mean, 
a great deal of trouble is taken in designing proper acoustics for reproducing a play 
in a theater, and proper acoustics for the recording setup. Now, if you have proper 
acoustics in the recording setup, and then you reproduce it in another space with 
proper acoustics, do not the acoustics of the first add to the second, and thereby in- 
crease the reverberation? 

MR. DUNBAR: Yes. The reverberations of the place in which you picked up 

386 DUNBAR Vol 49, No. 4 

the sound add to that of the new place in which it is put. The addition is not 
arithmetical but the confusion is. 

MR. JORDAN : On that basis, if the acoustics are designed for the reproducing 
end, why would not the proper acoustics in the recording end be with no rever- 
berations whatever? 

MR. DUNBAR: There are several reasons. One is, you do not know what room 
conditions you are going to have when you listen to it. The recorded programs 
might be given in any kind of room. A recording with no room effect is a dead 
staccato thing. Another reason is that musicians cannot work well in a dead room. 

MR. JORDAN: That is true, but it has been rrty observation that quite often 
a recording which has been made under what are thought to be ideal studio con- 
ditions, and played in a very modern theater which was supposed to have been 
designed by acoustical experts, sounds pretty terrible. 

MR. DUNBAR: A lot of theaters do reverberate for small audiences. There are 
many times when the amount of reverberation in a theater is a reflection of how 
much money has been spent to correct it. Exuberance might have led toward 
too much and of the wrong kind of materials, and then some other faults may 
show up. 

MR. DAVIS: Are there limits of temperatures and humidity within which 
acoustic measurements should be made? 

Would not there be a difference if the room were cold than if it were at normal 

MR. DUNBAR: It changes the high-frequency absorption to some extent. At 
3000 or 4000 cycles it would have a decided effect, but below that, I do not think 
it makes much difference whether it is hot or cold, or moist or dry. The low fre 
quencies seem to be little affected by it, unless the sound-absorbing material is 
saturated with water. 

MR. SELIG: Do pillars in a room have any effect on acoustics, and is there any 
way to treat them? 

MR. DUNBAR : Yes, they do. It depends on how large they are. The larger 
they are the lower the frequency that is scattered from the pillar surfaces. If the 
space in back of a row of large columns is highly reflective, you will have rever- 
berant feedback. If the columns are large, you get a scattering effect. In the 
Beaux Arts Building we used elliptical forms for coverings for the columns, and 
it seemed to work out very well; at least, the recordings sound pretty good. 

CHAIRMAN JAMES FRANK, JR. : Are there any new types of treatment particularly 
recommended for theaters in contrast to the types used during the first five or ten 
years after sound was introduced to theaters? Where is the best average place to 
put them? 

MR. DUNBAR: I think I have already touched on that. The most important 
place is the back wall. If you have many balconies, you have the absorption of 
the balcony areas, and you have less back wall to treat. 

The type of treatment has not changed a great deal. I think rock wool is now the 
basic material. Theaters nowadays have to use fireproof materials. Rock wool 
offers good absorption and is fairly inexpensive. The usual coverings for it are 
flameproof fabrics or glass cloth or perforated transite or metal. I am in favor of 
a material that reflects high frequencies, like perforated transite, or even thin light 
flameproof plywoods. 

Oct. 1947 SPACE ACOUSTICS 387 

MR. RADAMACHER : Where is there such a theater that has that type of treat- 

MR. DUNBAR: Eastman Kodak has a small projection studio theater that is 
treated that way, and it has been pronounced very satisfactory. Studio 8-H of the 
National Broadcasting Company and RadioCityMusicHallarealsogood examples. 

MR. RADAMACHER : How about acoustic plaster? 

MR. DUNBAR: That is a little indeterminate. Its value depends on the work- 
man who puts it on. If he presses it on too hard, it becomes dense and will absorb 
little sound. If it is applied properly it may fall off in places on account of poor 
bonding. It is not a fabricated material of fixed absorption, so its absorption may 
vary considerably. Materials of that type, when painted, lose a great deal of 

MR. GRIFFITH : I am very much interested in the historic shift of musicians from 
outside to inside. Are there any treatises on that which might be available? 

MR. DUNBAR: It is a theory of mine. I don't know whether there is anything 
published on it although there is a reference to it in the article on acoustics in the 
Encyclopedia Britannica, 

MR. LEWIN : Would you care to say what is the ideal reverberation value you 
can get ? Do you try to get it flat ? 

MR. DUNBAR: No, you usually try to get it fairly flat, but with the reverbera- 
tion tilted up a little at the low end and at the high end. The use and volume of 
room both determine the amount of reverberation and the shape of the curve. 

MR. FAY: Have there been any tests in recent years, by an unbiased source, 
of the absorptive values of different types of chairs manufactured by various 
theater-chair manufacturers? 

MR. DUNBAR: The National Bureau of Standards made some tests a number 
of years ago. There is also a new Acoustical Material Association booklet on 
absorption values which includes some information on chairs. 

MR. McGuiRE: Mr. Braun has been connected with the Radio City Music 
Hall since it opened. I wonder if he would like to make any comment on the 
subject we are discussing. 

MR. H. B. BRAUN: You asked before whether we have any trouble. I might 
touch on just one point. That is the use of acoustic plaster. In addition to the 
uncertainties because of the manner in which the material is. applied, there are 
also difficulties as a result of time. Although we have no absolute figures, we 
are quite certain that acoustic plaster does age and lose its effectiveness, and as 
a result, we have some echoes now which did not exist immediately after the 
construction of the theater. Our rear walls are very heavily treated, and some 
correction was applied subsequent to 'the completion of the building by means 
of changing the contours of some of our surfaces. 

MR. LEWIN: They have something over there like a public-address system. 
So far'as I know it is an unusual installation. Has it any effect on this problem? 

MR. BRAUN : It has this effect : because of the large size of the auditorium, 
which is nearly 2,000,000 cubic feet, it is necessary to reproduce at levels consider- 
ably in excess of unity. We are creating artificial reverberations by means of a 
reinforcing system. 

MR. LEWIN: Has the acoustic plaster ever been painted since it was installed? 


388 DUNBAR Vol 49, No. 4 

MR. BOYCE NEMEC: Would you describe the system of suspended absorptive 
cones that have been proposed for acoustic systems in rooms that can't be treated 
by normal wall or ceiling treatments, and give us some typical applications where 
a treatment of that kind would be suitable? 

MR. DUNBAR: They are functional sound absorbers developed by the Radio 
Corporation of America. The system is used primarily in places where the ceiling 
is so high or has so many pipes and obstructions that it is difficult to do anything 
else. They are cones, and two together make up a unit. They are hung up or 
strung up on stretched wires across the room. For a long time we have known 
that by breaking-up treatment from a concentrated area, it is more efficient per 
square foot. For instance, a single panel of 12 X 12 feet on the ceiling will have" 
a great deal more effective absorption if it is split up and each of the 144 square 
feet installed separately about the room. At the present time, the cones can- 
not be used in New York because they are nonfireproof . We are figuring on some 
means to make them fireproof. 

MR. NEMEC: Would you give us one or two examples of the type of rooms that 
can be treated with them ? 

MR. DUNBAR : One is a workshop where you have piping and belting, over the 
ceiling, and where you cannot put on so many square feet without excessive waste 
in cutting. Another is the case of an excessively high ceiling where, if you treated 
the ceiling, it would do very little good, but if you hang the material halfway 
down, you could increase the absorption. 

MR. BRAUN: Can you effectively reduce the volume by that method? 

MR. DUNBAR: Yes, and you get the absorption coming and going, if you hang 
enough of them. Generally they are hung pretty thick in a case of that kind. 

MR. GRAF: What is the material used for those cones? 

MR. DUNBAR: Very light wood fibers put together in a mat form, something 
like papier-mache. They are matted in shapes similar to a Mexican hat, and a 
pair of them together form a unit. When the sound energy flows toward it, instead 
of taking out just what strikes it, the sound waves passing by tend to buckle in 
behind and be absorbed the same way, so you pick up more sound energy than you 
do with the projected area of the same material applied flat on the walls. 

CHAIRMAN FRANK: Would you say that they could only be used in a place 
where decoration was not such an important factor? 

MR. DUNBAR: That would be my idea, although they could be used as part of a 

MR. GRAF: Are they usable over a rather wide frequency range so they can be 
used effectively in areas where other treatments would not be effective? 

MR. DUNBAR: They have a rather wide frequency range of absorption, but so 
have some other materials. It is not new from that standpoint but it is an eco- 
nomical form of obtaining absorption with a low-cost installation, where a nonfire- 
proof material would be acceptable and where other materials would be difficult 
to install. 


The newly incorporated Motion Picture Research Council, Inc., succeeds the 
former Research Council of the Academy of Motion Picture Arts and Sciences. 
Wallace V. Wolfe, active hi Society affairs for many years and current Chairman 
of the SMPE Pacific Coast Section, has been appointed Director of Research by 
the Board of the Association of Motion Picture Producers which will administrate 
and finance the Council's operations. 

Y. Frank Freeman, Chairman of the Board of the Association and also Chairman 
of the Research Council, announced this change and also reported that the 
Association recently appropriated $150,000 as an initial allotment to promote an 
industry-wide research program. 

The Council will expand its activities in all phases of motion picture develop- 
ment work, including: 

1. Designing and supervising construction of special equipment and processes 
for motion picture production. 

2. Applying new research developments to the industry and co-operating with 
universities and industrial research groups. 

3. Standardizing equipment and processes within the industry to permit 
better equipment to be manufactured more efficiently. 

The Motion Picture Research Council will be able to provide manufacturers 
with specifications on equipment which represent the desires of all the studios, and 
will welcome and act as a clearing house on all new ideas and information pertain- 
ing to motion picture production. 

William F. Kelley, Manager of the Academy Research Council for 11 years, 
will continue in his position as Manager and will take on additional responsibilities 
as assistant to Mr. Wolfe. The Council has engaged additional motion picture 
engineers and from time to time will assign others to specific projects. 

Present members of the Research Council, representing technical departments 
of each of the major studios, will assist the new staff. In addition to Freeman 
these members include Thomas T. Moulton, vice-chairman, 20th Century- Fox; 
John Aalberg, RKO-Radio; Daniel J. Bloomberg, Republic; Farciot Edouart, 
Paramount; Bernard Herzbrun, Universal-International; Nathan Levinson, 
Warner Bros.; Lohn Livadary, Columbia; Elmer Raguse, Hal Roach; Gordon 
Sawyer, Samuel Goldwyn; and Douglas Shearer, Metro-Goldwyn-Mayer. 

Design of new equipment will cover items not now available on the market, but 
the Council in supervising construction of working models by existing supply 
companies will not become a manufacturing organization. 

"The Council will analyze studio problems and work out solutions by combining 
the best ideas of the studios' own technical experts," Freeman said. "It will then 
serve as a co-ordinating agency to work with manufacturers to provide standard- 
ized equipment at economical prices. 

"We are fortunate in having Mr. Wolfe, with his outstanding record, to head our 
new program. The need for more thorough research into all aspects of film pro- 
duction has long been evident in Hollywood and other centers. We are confident 
that our new program will strengthen Hollywood's leadership in the field of mo- 
tion pictures.". 

Independent producers not now members of the AMPP will be extended the 
privilege of participating in the research work and its benefits, Freeman added. 




Because of the rapid growth of the Society of Motion Picture Engineers and the 
resultant increase in the office staff, it was found necessary to move Headquarters 
from the Hotel Pennsylvania in September. The new address is ninth floor, 
342 Madison Avenue, New York 17. 

It is hoped that our members will call upon us whenever they find it convenient. 
We extend to them a hearty welcome and hope that many of them will avail 
themselves of the opportunity to make us a visit. 

Executive Secretary 


It is planned, in the very near future, to publish material concerning our mem- 
bers, as well as information on new products which will appear in the market, 
Therefore, it is desired that members of the SMPE who make changes in position 
send to the Editor of the JOURNAL announcements concerning themselves. 
These write-ups should be phrased in a formal style, should be accompanied by a 
glossy print of the writer, and should not be*unduly commercial. Photographs 
should be of the formal, studio type, preferably 8 by ID inches in size. 

The subject matter should contain the following: date and place of birth; 
academic degrees granted ; dates and places of employment ; memberships in other 
scientific organizations. Necessarily, this invitation can be extended only to 
members of the Society, and the SMPE reserves the right to edit this material as 
it sees fit. 

New Products notes should be about items of outstanding interest to the mo- 
tion picture engineer. The commercial aspect should not be stressed, but the 
scientific value should be emphasized. Wherever possible, glossy prints of the 
article in question should be submitted with the write-up. The SMPE reserves 
the right to accept or reject material and to rewrite it if necessary. 


SOUND RECORDING ENGINEER: 16- or 35-mm equipment, single 
- or double system. Preferably educational or industrial films. Free to 
travel. For details write or phone Marvin B. Altman, 1185 Morris Ave- 
nue, New York 56, New York. Telephone Jerome 6-1883. 



Vol 49 NOVEMBER 1947 No. 5 



A Test Reel for Television Broadcast Stations 391 


The Showmanship Side of Television 


Design Factors in 35-Mm Intermittent Mechanisms 


Lightweight Recorders for 35- and 16-Mm Film 

M. E. COLLINS 415 

A Photoelectric Method for Determining Color Balance 
of 16-Mm Kodachrome Duplicating Printers 

PAUL S. AEX 425 

Portable and Semiportable Loudspeaker Systems for 
Reproducing 16-Mm Sound on Film 


A Survey, 8-Mm Problems ROBERT E. LEWIS 439 

Design Progress in an 8-Mm Projector 


The Movie-Sound-8 Projector LLOYD THOMPSON 463 

A New Sunshade and Filter Holder for 16- and 8-Mm 
Motion Picture Cameras JAMES T. STROHM 468 

Method and Equipment for Checking Motion Picture 
Apparatus Speeds C. T. OWLETT 471 

Current Literature 479 

Society Announcements 480 

Copyrighted, 1947, by the Society of Motion Picture Engineers, Inc. Permission to republish 
material from the JOURNAL must be obtained in writing from the General Office of the Society. 
The Society is not responsible for statements of authors or contributors. 

Indexes to the semiannual volumes of the JOURNAL are published in the June and December 
issues. The contents are also indexed in the Industrial Arts Index available in public libraries. 




342 MADISON AVE. NEW YORK 17, N. Y. TEL. MU 2-2185 







** President: LOREN L. RYDER, 

5451 Marathon St., Hollywood 38. 
** Past-President: DONALD E. HYNDMAN, 

342 Madison Ave., New York 17. 
** Executive Vice-President: EARL I. SPONABLE, 

460 West 54th St., New York 19. 
* Engineering Vice-President: JOHN A. MAURER, 

37-01 31st St., Long Island City 1, N. Y. 
** Editorial Vice-President: CLYDE R. KEITH, 

233 Broadway, New York 7. 
* Financial Vice-President: M. RICHARD BOYER, 

E. I. du Pont de Nemours & Co., Parlin, N. J. 
** Convention Vice-P resident: WILLIAM C. KUNZMANN, 

Box 6087, Cleveland 1, Ohio. 
** Secretary: G. T. LORANCE, 

63 Bedford Rd., Pleasantville, N. Y. 
* Treasurer: E. A. BERTRAM, 

850 Tenth Ave., New York 19. 

**JOHN W. BOYLE, 1207 N. Mansfield Ave., Hollywood 38. 

*FRANK E. CARLSON, Nela Park, Cleveland 12, Ohio. 

*ALAN W. COOK, Binghamton, N. Y. 
**ROBERT M. CORBIN, 343 State St., Rochester 4, N. Y. 
**CHARLES R. DAILY, 5451 Marathon St., Hollywood 38. 
*tjAMES FRANK, JR., 356 West 44th St., New York 18. 

*JOHN G. FRAYNE, 6601 Romaine St., Hollywood 38. 
**DAVID B. JOY, 30 East 42d St., New York 17. 

*PAUL J. LARSEN, 1401 Sheridan St., Washington 11, D. C. 

*WESLEY C. MILLER, MGM, Culver City, Calif. 
**HOLLIS W. MOYSE, 6656 Santa Monica Blvd., Hollywood. 
*JA. SHAPIRO, 2835 N. Western Ave., Chicago 18, 111. 
*WALLACE V. WOLFE, 1016 N. Sycamore St., Hollywood. 

Term expires December 31, 1947. tChairman, Atlantic Coast Section. 
Term expires December 31, 1948. tChairman, Midwest Section. 
Chairman, Pacific Coast Section. 

Subscription to nonmembers, $10.00 per annum; to members, $6.25 per annum, included in 

their annual membership dues; single copies, $1.25. Order from the Society at address abort. 

A discount of ten per cent is allowed to accredited agencies on orders for subscriptions 

and single copies. 
Published monthly at Easton, Pa., by the Society of Motion Picture Engineers, Inc. 

Publication Office, 20th & Northampton Sts., Easton, Pa. 

General and Editorial Office, 342 Madison Ave., New York 17, N. Y. 

Entered as second-class matter January 15, 1930, at the Post Office at Easton, Pa., 

under the Act of March 3, 1879. 


Vol 49 NOVEMBER 1947 No. 5 


M. R. BOYER** 

Summary. After rechecking the work of the Subcommittee on Film of the Tele- 
vision Committee, it was found that broadcast stations needed a test reel which would 
allow dynamic checks of their systems. A reel was manufactured and its possible 
uses are described. 

In the years since the publication of the report by the Subcommit- 
tee on Film Tests of the SMPE Television Committee 1 much progress 
has been made in television broadcasting. It was felt that some of 
the work of this committee would bear repetition on the improved 
systems in order to check previous results. 

A suitable reel was secured and four prints were made using nor- 
mal- and low-contrast stock, and printing at various light points. 
These were then projected, transmitted, and viewed on the moni- 
tors of four different stations. In general our results were as follows : 
A print two points lighter than normal televised better than a normal 
print; and the less brightness range in the original subject the better 
the details were reproduced in the high lights and the shadows. 
Where no automatic bias control was used, a flat print projected bet- 
ter than a print of normal contrast. This agrees quite well with the 
previous report, although in this case all tests were made using an 
iconoscope as a pickup tube. 

Again the above conclusions conform with theory in that the lighter 
print allows more light to fall on the mosaic plate and so requires less 
amplification. Also, low-brightness range in the original scene or 

* Presented Oct. 16, 1945, at the SMPE Convention in New York. 
'* E. I. du Pont de Nemours and Company, Photo Products Department, 
Parlin, N. J. 


392 BOYER Vol 49, No. 5 

print means that the picture-tube screen, capable of limited-bright- 
ness range, can more accurately reproduce the original without losing 
details in either the high lights or shadows. 

In the course of these tests two things were noted. The system per- 
formance was usually checked prior to broadcasting by projecting 
with a slide projector, a still pattern onto the mosaic. No dynamic 
test of the system was made prior to going on the air. Second, no 
method was currently used to check the over-all tone reproduction of 
the system when projecting film. All checks had been made with 
static test patterns. 

In discussing these two observations with station engineers it was 
learned that a film containing one or more somewhat standard test 
patterns would be welcome. At the same time it was decided that a 
step-density wedge which could be photographed from the monitor 
tube would allow a dynamic check of the whole system, should a sta- 
tion desire to make such a check. These two features, plus various 
types of scenes printed at different light points, apparently would 
comprise a most welcome aid to station engineers. 

Accordingly work was started in assembling such a reel. It was 
found that each station had its own preferred test pattern, in addition 
to those that had been worked out by the Radio Corporation of 
America for checking their iconoscopes and kinescopes and which 
had been published in the literature. Also, after running the partly 
assembled reel at several stations, other suggestions were made for 
types of patterns to be included. One of these was the superimposi- 
tion of dots with varying densities on each step of the wedge. In 
making the titles, it was felt that it would be helpful if one title had 
black letters on a white background, and the following title had white 
letters on a black background. This would indicate the time required 
for shading successive scenes where each scene was drastically different. 

After frequent checking with four major stations, the new reel was 
assembled. It must be realized that many test patterns had to be 
omitted, and those used by three stations were included as being 

The reel is composed of the following scenes and titles. A descrip- 
tion of the chart and what may be noted in transmission defects is 

Titles 1 and 2 are for use in checking the rapidity with which shad- 
ing can be accomplished when going from a clear background to an 
opaque background. 


The NBC chart is so designed that the ratio of the radius of the 
large circle to the inner circle is 4 to 3 or the prescribed aspect ratio. 
If the circles have a true form the aspect ratio is correct. However, if 
the circles have an elliptical shape the pattern is too wide when the 
main axis is horizontal, and too narrow when the axis of the ellipse is 
vertical. Nonlinearity in ratio of scanning is indicated by an egg- 
shaped outline. 

The "eye" of the pattern is divided into four rings of equal density 
differences. Therefore for true reproduction of high lights and 
shadows the apparent difference in brightness between adjacent rings 
should be the same. 

Subject Footage Running Time 

1 . Title Clear Letters on Black Background 30 ft 20 sec 

2 . Title Black Letters on Clear Background 30 ft 20 sec 

3. NBC Wedge Chart 90ft 1 min 

4. DuMont Wedge Chart 100 ft 1 min 

5. Philco Wedge Chart 100ft 1 min 

6. Checkerboard Chart RCA 100ft 1 min 

7 . Step Densities 100 ft 1 min 

8. Step Densities with AD 100ft 1 min 

9. 5 Scenes Normal 90ft 1 min 

10. 5 Scenes 2 Light Points under Normal 90ft 1 min 

11. Last Scene 4 Light Points under Normal 6ft 4 sec 

12. Last Scene 6 Light Points under Normal 6 ft 4 sec 

13. Half and Half ; 60ft 40 sec 

The RCA chart is designed so that the squares are duplicated in 
sections allowing the defects to be localized. The shape of the main 
square and subordinate squares indicate the geometrical properties 
of the image. Because the chart is designed with a height of three 
squares and a width of four squares, the aspect ratio is easily checked. 
Any nonlinearity in either direction will be shown by a change in 
shape of the squares. 

1234 56 

Density: 0.23 0.48 0.92 1.50 1.77 1.80 

Transmission per cent: 0.59 0.33 0.12 3.2 1.7 1.6 

Any pairing in interlaced patterns will be shown by an uneven ap- 
pearance in the horizontal wedges at the point of highest resolution. 

Density-step tablets are inserted so that an over-all check of the 
contrast-transmission characteristics of the system can be made by 
direct photography of the tube screen. 

394 BOYER 

Step densities with A> is a repetition of the step-density tablet but 
has dots of density 0.1, 0.2, and 0.3 superimposed on each of the six 
densities. This chart shows the minimum density differences which 
can be picked up at various locations along the tonal scale. 

Five scenes, printed "normal" are a series of typical scenes from 
du Font's "Soldiers of the Soil" printed at normal print density for 
theater use. 

Five scenes, printed two points light represent an average density 
0.04 less than normal and transmit 10 per cent more light. 

Last scene printed four points light is printed four points lighter 
than the normal print. It has an average density of 0.08 less than 
normal and transmits 20 per cent more light. 

Last scene printed six points light is again printed light but this 
time six points lighter than normal. It has an average density 0.12 
less than normal and transmits 32 per cent more light. 

The gradations in the sky areas of this scene form a practical ap- 
plication of the second density tablet. 

Half field clear and half opaque and reverse is to check ability to 
keep sharp demarcation, or pairing. It also indicates the memory of 
the iconoscope mosaic. 


We should like to acknowledge the kind co-operation of National 
Broadcasting Company, Columbia Broadcasting System, DuMont 
Laboratories, Inc., Philco Radio and Television Corporation, and 
General Electric Company in furnishing either material or sugges- 
tions, or both, in the preparation of this television test film. 


1 Report of the Sub Committee on Film Tests SMPE, 35, 6 (Dec. 1940), 
pp. 580-583. 


Summary. The possibilities of theater television are discussed strictly from 
the box-office viewpoint. There is also a compilation of the many events which 
would make good box-office fare, and a description of many new uses to 
which theater television may be put which would attract people to the theater. 
How theater television may supplant the present motion picture programs and a posi- 
tive viewpoint on theater television as a legitimate and necessary adjunct to the motion 
picture theater are also considered. 

During the twenty-six years I have been associated with the radio, 
motion picture, and television industries I have developed great 
admiration and profound respect for the engineer. I can recall no 
instance so far where in any of these arts the research or the practical 
engineer has failed to deliver first a technically workable, and then a 
commercially feasible instrumentality. Of course, you always hear 
the calamity howlers dismissing a new invention by wailing: "It 
can't be done" "It's no good" "It's a novelty" "It won't last"- 
and so on. But there's a homely proverb which warns: "During the 
time you spend listening to the reasons why a thing can't be done, 
some one else is doing it." 

I feel I can safely make the following assumption. The engineering 
fraternity is going to produce a technically feasible theater television 
system for the motion picture and television industry. The question 
is, what are we commercial men, the producers, the showmen, the 
merchandisers, the exhibitors, going to do with this technically perfect 
system? I, for one, am very much in favor of theater television. I 
have not heard many opinions one way or another from nontechnical 
people in the field of motion pictures. But I did run across a paper 
on theater television written by an engineer a practical engineer, 
too which set out to prove that, in the final analysis, television had no 
place in the theater because it could not deliver to a theater audience 
a program which was high enough in showmanship to make the show 
pay. This paper surprised me tremendously, largely because for the 
first time in my memory, here was a defection in the ranks of the 
engineers. Here was one of our own number vehemently stating that 

* Presented Oct. 24, 1946, at the SMPE Convention in Hollywood. 
** President, RKO Television Corporation, New York, N. Y. 


396 AUSTRIAN Vol 49, No. 5 

theater television would probably never be commercially possible. 
This is a very tender subject for an engineer to comment upon. As 
a matter of fact, in the past, questions of this nature have proved 
themselves to be very tender subjects for some of the greatest names 
in the motion picture industry. 

Let us go back some twenty-one years ago when the engineers who 
had developed a device for making talking motion pictures tried to 
convince the producers of silent motion pictures that they had a sal- 
able commodity. The record contains some very interesting facts. 
Let me read you some of the utterances of men who were and still 
are big names in the motion picture industry. All but one of them 
whom I shall quote are alive and active in business today, so for that 
reason, I shall not mention any names. 

The head of a company said : 

"I do not know what 1929 will bring to the industry and neither does anybody 
else. Last year brought talking pictures. No one could have predicted what a 
rage they would become. In my opinion they will be just as much a rage next 
year. But I think that as we get further into them, we shall find that eventually 
talking pictures will be restricted to certain types of productions rather than that 
the trend will be to make all pictures talk. The silent picture technique is too well 
established and its popularity too widespread to permit a variation of its form, no 
matter how interesting it is to the public, to dominate the thing itself." 

A great theater chain head, said : 

"Whether the major feature picture success of 1929 will be all-talking or part- 
talking or synchronized or silent, I believe it is impossible for anyone to say at 
this moment. Certainly the combination of the advantage of the silent picture in 
its swift action, its pantomimic advantages, and its photographic beauty, com- 
bined with the proper musical setting and augmented in its climaxes with spoken 
dialog would seem at the present minute to be the outstanding picture success 
of the coming year, always provided, however, that the story is worth the telling. 
With the prosperity that is in the country today I look for continued prosperity 
in the picture theater, if production measures up to the standards which will be 
demanded of it." 

(This was a masterpiece of carefully calculated indecision.) 
A sales manager said : 

"The introduction of sound and talking pictures has created a new clientele at 
the box office and will draw to the motion picture theater this coming year mil- 
lions of people who heretofore have never patronized theaters to any great 


extent. This will reflect in a healthy condition to both the producer and distribu- 
tor. It is my opinion that no theater in a city of any size can exist without pre- 
senting some part of its program in sound or dialog. However, I do not wish 
you to construe this statement to the effect that silent pictures should be elimi- 
nated. I contend that a fine silent picture properly synchronized to music will 
draw as well as any talking picture." 

The head of a great studio said : 

"A passing fad," said Mr. X, of the talkies. He added that "they might be all 
very well for the moment, but that the traditional silence of the silent drama 
would never die. When the public had had time to become fed up with this 
mechanical novelty, it would rally back to the old standards of subtitles and 
speechlessness. " 

A very well-known writer both for radio, pictures, and television 
wrote a magazine article in November, 1929, entitled "The Movies 
Commit Suicide". I quote herewith a few of his more precious ob- 
servations : 

"After some twenty years of being only in its infancy, the moving picture which 
gave promise of an interesting adult life, has again suddenly become senile and 
garrulous" .... "The effect of the new movies on the stage will depend largely 
on its effect on the old movie. According to the enthusiasts, the silent movie is 
doomed. I should say that in that case, the stage, although it has nothing what- 
ever in common with the silent movie, will also go under. If the talking movie 
can undermine one, it can undermine the other" .... "The alternative which I 
think very likely to happen is that the film with just dialog will become a separate 
form of entertainment drawing to itself nearly everything tawdry and vulgar in 
the silent film and leaving the silent film in the hands of people, mostly foreign 
and amateurs able to appreciate its values" .... "It is a little too early to de- 
clare that the talking movie picture will become a cesspool for both the movie and 
the stage. I think that this is likely to happen because the movies are still in 
the hands of producers and directors who seem never to have taken the slightest 
interest in their own medium and have never studied its resources, its mechanism 
and its technique and its effect." 

What an indictment this was. The jury, however, brought in the 
verdict of "not guilty". 

A great director-producer said : 

"Speech will be used by those directors who do not understand the capacities 
of the moving picture, whereas those who do, will extend their mastery over their 
own instrument instead of calling in an alien element." 

A director-producer said : 

"Sound pictures, that is, with dialog that runs continuously will do away en- 
tirely with the art of motion pictures." 

398 AUSTRIAN Vol 49, No. 5 

Another director said : 

"You can't convince me that continuous dialog in a picture will do anything but 
detract from its value. ' ' 

One of the greatest said: 

"I do not believe there is anything revolutionary about the advent of talking 

(It is generally conceded that the advent of sound caused one of the 
world's greatest revolutionary technological upsets.) 

A headline in the Motion Picture News where, incidentally, prac- 
tically all of these quotations appeared says : 

"Hollywood is not excited over talkies." 

A very prominent chain exhibitor said : 

"However, I will never believe that the mechanical reproduction 
of synchronized music or voice will ever be of any greater value to the industry 
than just the novelty of it at the beginning." 

Our own beloved Thomas A. Edison, I take the liberty of using his 
name and with the utmost respect and reverence, said: 
"I don't think the talking picture will ever be successful in the United States." 

And last but not least here's an Editorial from the Motion Picture 
News of June 2, 1928: 

"Furthermore, it is a well-known fact that it is a hard matter to entice members 
of the New York colony away from Broadway, particularly when they are enjoy- 
ing favorable salaries." 

It makes quite a record. Believe me, please, when I say that I have 
no desire to ridicule the authors of these statements. They have 
proved themselves to be keen businessmen and master showmen 
many times over. The only trouble with their thinking 21 years ago 
was that the public did not agree with them. They thought they 
knew what the public wanted but for once, they missed. To their 
eternal and everlasting credit, they later reversed themselves, and 

But after it was all over, and everyone had his say, and the Brothers 
Warner showed them all the way, the editor of the Motion Picture 
News released the following significant statement: 

"What after all started this rage that now has the large motion picture indus- 
try by the ears? Just one picture, gentlemen, The Jazz Singer. One box-office 
picture did the trick. That should be borne in mind." 


And the editor was right. This fact should indeed be borne in 
mind, gentlemen. Because the same thing is going to happen with 
theater television. It's only going to take one sellout to start the ball 
rolling. It's only going to take the vision and the courage of one man 
or one company to start theater television on its way precisely as one 
company started sound pictures on their way despite the practically 
unanimous opinion of the industry that it was a foolhardy, silly 
novelty. There was a wonderful celebration of the Sound Picture's 
Twentieth Anniversary in August, 1946. Perhaps that in itself 
should stand as proof that engineers should stick to engineering. I 
feel sure that if they do, they will always find a group of individuals 
with enough foresight, imagination, and money to make good use of 
their inventions. 

I am not worried about what any engineer or group of engineers has 
to say about the commercial future of theater television. Nor am I 
going to become involved in a discussion of what is the best kind of 
theater television equipment; direct television which appears on the 
screen of a theater while the event is actually happening or delayed 
theater television which is accomplished by practically instantaneous 
recording on film of both the image and the sound. 

But, I am worried about the men who have charge of putting enter- 
tainment of one kind or another inside the motion picture theaters of 
America. English showmen seem to be much more alert to the possi- 
bilities of theater television. In 1946 there was present in this 
country a delegation of six of vhe executive staff of one of the large 
British picture companies who also operate many theaters, for the 
purpose of finding out all we know about theater television. So long 
as I am connected with the American motion picture industry, it is 
my aim to keep the subject of theater television constantly in the 
minds of the American theater owner. I think I can prove that 
there are enough events of public interest to make theater television 
a paying proposition to the theater owner. Here is a pretty accurate 
way of forecasting the answer to this question. 

Let us, for instance, take one single, one time-a-year event such as 
the Kentucky Derby. The racetrack at Churchill Downs in Louis- 
ville has a very small capacity, about 65,000. The "sport of kings", 
however, has a tremendous following scattered throughout the length 
and breadth of this fair land. Any horseflesh fancier who has ever 
laid a $2.00 bet on the nose of some "nag" would jump at the oppor- 
tunity to see the running of the Kentucky Derby. The exhibitors of 

400 AUSTRIAN Vol 49, No. 5 

America showmen at heart will not be slow to visualize this tre- 
mendous potential box office. Their programming agency would, I 
am sure, be able to consummate a deal with the Churchill Downs 
authorities under whose auspices the race is held whereby, for the pay- 
ment of a rather substantial sum of money, this event would be tele- 
cast exclusively to the theaters of America. In these theaters there 
are approximately 11,700,000 seats. I daresay that the privilege of 
witnessing the Derby not from a seat somewhere behind a post, or 
from the infield without a seat, but from a comfortable chair in one's 
own neighborhood theater for, let us say, one or even two dollars, 
would be eagerly accepted. 

It would not be a bad seat either, for you can rest assured that the 
television cameras will be so placed that millions of pairs of eyes in the 
theaters of America would have a ' 'down-front" seat. As a matter of 
fact they would have better than a down-front seat. There would 
undoubtedly be a television camera stationed at each furlong post 
and the millions of watchers literally would be going around the track 
with the thoroughbreds. Watching from a theater seat would be in- 
finitely better than from a clubhouse seat at the track. You would 
hear the frenzied excitement of the crowd, the thundering of hoof 
beats. You would actually be there without leaving your home 
town. I feel certain that the Churchill Downs people would be in- 
clined to make this kind of a deal, and I am sure that no sponsor of 
telecast programs could afford to meet the ante of the exhibitor. 
This is a roundabout way of my saying that the event would be 
shown in the theaters only and would not be telecast for home con- 

Here are a few other events of national interest which would make 
excellent theater television fare: championship boxing matches, 
critical major-league end-of -season baseball games, the world series, 
men's national tennis championships, women's national tennis 
championships, all important intersectional and many local college 
and professional football games, basketball games, indoor athletic 
games, finals of the national horse show, Westminister kennel club 
finals, national open golf championships, and national bowling 

I could go on and on and on. But even then not all the possibilities 
would be covered, for as the art moves on we shall find additional 
events possessed of drawing power which do not enter our minds at 
present. For instance, twenty years ago you would never have 


believed that one of radio's most outstanding coast-to-coast stars 
could be a wooden ventriloquist's dummy. But it is surprising how 
many people I run into now who said they thought so. Hindsight is 
a very comfortable facesaver. 

Objection has been raised to the effect that the inclusion of a theater 
television program as part of the regular fare of a motion picture 
theater will interfere with the proper scheduling of a show. I do not 
fear this. Ways and means will be found to include any event worthy 
of showing in the theater by rescheduling the show. There need not 
be much, if any, breaking into the middle of a feature picture in order 
to show some local happening. That type of event could probably 
be shown at the end of the picture if the so-called storage method of 
television were adopted. Nor do I hold with the objection that 
theaters do not have room for a television projector. Room will be 
found if they need a hole in the floor they will cut one; if they need a 
larger booth, they will build it. 

Then there are those who raise questions regarding possible labor 
disputes of the jurisdictional type. The picture industry is used to 
those things. Theater television will make it possible for the theater 
owner to make his theater the amusement and cultural center of the 
neighborhood. Think of what that means. Under one roof your 
patron can have movies, music, current happenings, radio, sporting 
events, plays, personal appearances, and other interesting programs. 
Above all it will obviate the necessity of people staying home next to 
their television sets for fear they will miss something. Some of these 
events will be charged for, some will be thrown in as part of the show. 
The main thing is to keep them coming to the theater, the social center 
of the community. 

The purpose of this paper is "let the engineers engineer, and let the 
showmen worry about the show". Let each man stay within his own 
calling. Let the engineer believe in his creations and the showman 
think to a constructive future. I promise you that I will do all I can 
to keep the showman's mind focused on theater television with a 
positive attitude. And I shall do everything I can to make it im- 
possible for some future researcher to go to the files and dig out the 
funny articles and the negative opinions about theater television with 
which to address the 80th Semiannual Conference of the Society of 
Motion Picture Engineers. 

402 AUSTRIAN Vol 49. No. 5 


MR. ZIMMERMAN : Mr. Austrian makes a definite point that theater television 
of special events will be exclusively for theater owners, because their super eco- 
nomic powers can outbid a radio sponsor. Am I correct in that point? 

MR. R. B. AUSTRIAN: Yes, I laid the accent on economic power. 

MR. ZIMMERMAN: Right now radio and the motion pictures co-operate to a 
great extent. What will be the economic effect in the motion picture industry, if 
it begins to buck the broadcast television industry, which is at present, I think 
we can say, at least a 75 per cent radio operation, rather than a motion picture 
operation? Are you aware of the implications, or have you gone into that at all 
in making that particular statement? It is not meant as a challenge, I am just 

MR. AUSTRIAN: Yes, I have given some thought to that. I think that the 
motion picture producer or exhibitor, paradoxical as it may seem, will be one of 
the greatest customers of the networks. We are going to bring much time on tele- 
vision to put trailers of our pictures in your homes, so that the seesaw will be kept 
pretty accurate. We may compete with them on events, but also we shall buy a 
great deal of time. Is that along the lines you were thinking? 

MR. ZIMMERMAN : That is one answer. I am just interested whether the large 
broadcasters feel the pressure of the motion picture people. I think opinion only 
will be on special events, that is for television broadcasting, as well as theater 

MR. AUSTRIAN: I think that is right. I have often said without these spon- 
taneous events, games, parades, and so forth, who would want to buy a television 

MR. L. L. RYDER: I wonder if it would not be well to point out that the air and 
right to air are not necessarily held entirely by the broadcaster. The broadcaster 
is not necessarily the televisor. The broadcaster of today is the broadcaster of 
voice and not necessarily television, because it is a new field we are talking about 
rather than a competition with an old field. 

MR. AUSTRIAN : I think that is true, Mr. Ryder, and our Society has been very 
forward-looking in having assigned to it certain bands in the spectrum for theater 
transmission and relay, intercity and intracity. The Telephone Company to- 
day has certain bands over which people may speak privately from point 
to point, or from one point to many points. We consider theater television 
not broadcasting but a multiple-message service, whether it comes through the 
air or cables, we do not yet know. All we know is, it is coming. 

MR. LANSBERG : First of all, just as an example, we shall transmit football 
games not only to the home, but also to the theater. That will be a small re- 
ceiver but we hope to have a larger one available for screen projection in the not 
too distant future. I believe, also, in the serious work that has been carried on 
by Paramount Pictures in New York on two different methods of theater tele- 
vision. You can be assured that you are not alone. 

MR. AUSTRIAN : Bear in mind, I did not say no one else was doing it, I just 
said no one else was talking about it. 

MR. J. D. BRADLEY: I wonder if you televise through the air, if you can 
buy an exclusive on an event, if you would be allowed to under the Federal 


Communications Rulings? That is the first part of my question. The second 
part is, it seems to me you are going to have the giants spinning against each 
other, one to make an exclusive for the theaters, and another to put it in the homes 
on sponsored programs. It seems to me the attitude of trying to get an exclusive 
for the theaters is contrary to the present trend of trying to get it in the homes, and 
represents, perhaps, a justified but selfish attitude. 

MR. AUSTRIAN : Well, part one of the question : I do not know what the philos- 
ophy of licensing might be at a later date, especially if we approach it from 
multiple-message service rather than broadcasting. That can change. I do 
not think we do know the answer. I am looking at it now from possibilities. 
As to the second part of your question, of course, it will be as everything else is in 
this world, I guess, in business, a struggle between two people after the same 
thing. If the man who manufactures tooth paste can pay more for the Ken- 
tucky Derby than 17,000 theaters, he will get it, if he can find a way of recouping 
what he pays for it. However, he does not have a box office and the theaters' 
financial resources. If what is inside the theater interests you, you pay; and if 
what is inside does not interest you, you will not pay. So far, the theater men 
have been very active in putting on something for which a hundred million people 
ninety-five million people a week pay. So there are some large figures there, 
and if they can outbid the theater and get it, of course, they will. 

Miss CATHERINE SIDNEY: Will it be in color or will it be in black and white in 
the theater? 

MR. AUSTRIAN: You are speaking of television I am sure it will come in 
black and white first; and I am sure, also, that someday it will also be in 

MR. JAMES FRANK: I think it should be a matter of record that the Television 
Practice Committee of this Society has made quite a study of the development as 
far as it has gone at the present time of theater television equipment. At the 
present time the efforts of the Committee are at a standstill because we are 
awaiting a group in the industry to tell us what they want in theater television, 
how they will use it, and where, and when ; and up to the present time I believe 
it has been impossible to obtain any valid cross section of opinion from the ex- 
hibition phase of our industry. So, the Committee and the manufacturers and 
broadcasters and others represented on that Committee, are to some extent handi- 
capped in their efforts to develop further and to introduce commercially theater 

MR. AUSTRIAN: Frankly, I do not know as we know what we want. All we 
know is we want a good, big, bright picture. Now it is up to you to give it to us. 
Where it comes from, I do not think is too important. Some people expressed a 
horror at the idea of taking eight or ten or twelve feet out of the front of the bal- 
cony. Well, I think you could figure it out with a pencil and paper, if you figured 
the rate of earning per foot per month, or per year, and plot it against the rate of 
occupancy of the house, and figure that you can increase your rate of occupancy. 
Suppose you lose a few seats? You would still take in more dollars. As I said 
before, if they need a hole in the floor to put in this equipment, they will make it. 
The main object is, get a machine that will throw an acceptable, clear, bright, large 


DR. E. W. KELLOGG: How large a factor do you think it is in the enjoyment 
of a television program, the fact that you see it absolutely simultaneously with the 
action versus a few minutes' delay? 

MR. AUSTRIAN: I can only give you a personal opinion on that as a theater 
man. I think I should welcome a device upon which I would capture the image 
and show it to five shows later on in the day, and have that repeat proposition. 
If you take it as it conies, it is gone forever; but if you can photograph it and 
perhaps edit it, if you have a few minutes' time you can certainly run it five times, 
ten times, twenty times, before the newsreels will catch up with you. I think 
that is very important. So far as its being a few seconds behind, take this audi- 
ence tonight, for instance we are all in this room and perhaps there is a football 
game going on. It is over 30 seconds or 1 minute before we know it. It is news 
to us. We are isolated. It is just as if it happened. I do not think that would 
be too great a disadvantage. Also, you can cut in the event at a break in your 
program between a feature and a short, or between a short and the overture. 
You get great flexibility and a repeat proposition. 

MR. ZIMMERMAN : I have just finished an analysis of replies from all of the thea- 
ters of our Army and Navy personnel who have been listening to the playoff 
games of the National League and the World Series, and we have received a ratio 
of 5 to 1 of complaints because we rebroadcast the World Series instead of broad- 
casting it live, even though it was 3:00 o'clock in the morning in Tokyo when 
they could have received the live broadcast and we rebroadcast it at the earliest 
possible moment they would be up, namely, 7:00 o'clock in the morning. I think 
people want to know about events which have a score as to one side or the other 
as soon as they happen, rather than to wait for some little time. A few seconds 
would not make any difference. 
- MR. AUSTRIAN: How do you rebroadcast? 

MR. ZIMMERMAN: We rebroadcast them from the network lines. This was 
purely voice broadcast. 

MR. W. H. OFFENHAUSER, JR: Possibly an explanation of that might be that 
some of the boys had a bet on the game and wanted to pay off their bets as soon 
as possible. 

MR. AUSTRIAN : Some one has already said that theaters could open up pari- 
mutuel booking offices. 

MR. JOHN CRABTREE (dictated to reporter later as a question Mr. Crabtree 
would like to have printed in the JOURNAL, attention of Mr. Austrian) : The chief 
objection to present-day television images would appear to be lack of sharpness of 
the sirreen image. Admittedly in the early days of the sound picture, the sound 
quality was relatively poor, but the ear will tolerate a wide range in sound quality; 
but since the eye has been seeing nothing but sharp images since birth, it resents 
even the slightest lack of definition or sharpness. In my humble opinion, tele- 
vision will not receive widespread adoption until sharper images are available. 



Summary. The operation of a conventional Geneva-type intermittent mechanism 
is analyzed and used as a basis for developing several thoughts which may lead to a re- 
duction of pull-down time. Reduction of pull-down time is desirable in that it per- 
mits the use of shutter blades of smaller width, thus either increasing the screen bright- 
ness or permitting the same screen brightness with less power in the lamp. The in- 
crease in the acceleration of the film due to the reduction in the pull-down time natu- 
rally will increase the wear and damage to the film. Means are shown, however, 
whereby the pull-down time may be reduced without subjecting the film to any greater 
acceleration than occurs with the conventional Geneva-type intermittent mechanism 
and consequently will not increase the wear on the film. 

Intermittent mechanisms for moving the film, frame by frame, past 
the projection aperture have been used quite satisfactorily through 
all the years that motion pictures have been practical. Intermittent 
mechanisms of the Geneva type are now almost universally used in 
35-mm projectors. Other types such as the "Powers" have been 
used and are worthy of consideration. 

This is not being written primarily to describe things which exist 
but to describe a train of thought which might be followed in an at- 
tempt to devise a new intermittent mechanism. 

A starting point is always needed, so let us start with an analysis of 
the Geneva movement, inasmuch as that has been in wide use for a 
number of years, after which we can develop the above-mentioned 
line of thought regarding an intermittent movement which would re- 
duce the time of pull-down, increase the efficiency with which we put 
light on the screen and yet, at the same time, handle the film as easily 
as at present so as not to reduce the life of the film. 

In most projectors the length of the film which is subject to inter- 
mittent action is approximately 8 inches. This is the distance from 
the center of the intermittent sprocket to about 2 inches above the 
film gate. The weight of an 8-inch length of 35-mm film is ap- 
proximately 0.0032 pound, so that the mass of the film is therefore 

* Presented Apr. 25, 1947, at the SMPE Convention in Chicago. 
** General Precision Laboratory, Inc., Pleasantville, N. Y. 




Vol 49, No. 5 

approximately 0.0001 slug.* This is the first constant which affects 
the reasoning described herein and will be considered as having that 
value in all that follows. Fig. 1A indicates the above-mentioned 
length of film in relation to the aperture and the intermittent sprocket. 


Fig. IB indicates the principal relationships involved in the con- 
ventional Geneva intermittent mechanism. The radius of the circle 
on which the pin moves is taken as equal to one. The distance from 
the point of tangency as the pin enters the slot to the center of the 


FIG. 1. A, Film trap showing approximate length of 
film subject to intermittent action. B, Geneva intermit- 

star wheel is also equal to one. This makes the distance between the 
centers of the pin shaft and the star wheel equal to \/2. In connec- 
tion with this figure, we can develop expressions for the angular dis- 
placement or travel of the star wheel, the velocity of the star wheel 
and the acceleration of the star wheel, all as functions of the angular 
displacement or travel of the pin. 

1. Angular Displacement or Travel of Star Wheel. Let repre- 
sent" the angular displacement or travel of the pin and let it be 
measured from the line connecting the centers of the pin shaft and 
the star wheel. Let /3 represent the angular travel or displacement of 
the star wheel, this angle being measured from the same line. 


sin 6 

/3 = 


* NOTE: Mass in slugs = weight in pounds/acceleration of gravity in feet per 
second squared. One slug = approximately 32 pounds. 


Equation (1) above represents the displacement or travel of the 
star wheel as a function of the displacement or travel of the pin, both 
measured from the mid-point of the pull-down period. 

2. Velocity. Let u p represent the angular velocity of the pin 
and oj s represent the angular velocity of the star. Since velocity is 
the derivative of displacement with respect to time, 


* (2) 

= ^ 

1 = (4\ 



-y/2 COS 01 

P 3-2V2cos0 

Equation (5) is, therefore, the desired equation for the angular ve- 
locity of the star as a function of the angular travel or displacement of 
the pin assuming that the angular velocity of the pin is known. 

3. Acceleration. Let a s represent the acceleration of the star 
wheel. Since acceleration is the derivative of velocity with respect 
to time, 

da) t 


-, = * 0) 


j/\ * V * / 

COp (lO 

Solving this for a, we have 

a 9 = o, p -^ (3) 


\/2 sin 

* 7= (9) 

(3 - 2\/2cos0) 2 

408 HAYEK Vol 49, No. 5 

Equation (9) is the equation for acceleration of the star wheel as a 
function of the angular travel or displacement of the pin assuming 
that the angular velocity of the pin is known. 

Equations (1), (5), and (9) show the basic behavior of the Geneva 
intermittent mechanism itself. To obtain information regarding the 
film which is controlled or moved by the intermittent mechanism let 
us express /?, o> s , and a s in radians and the radius R of the intermittent 
sprocket in inches. Let 5 represent the travel or displacement of the 
film in inches per second and let a represent the acceleration of the 
films in inches per second per second. 


* = Rp (10) 

v = Rw t (11) 


a = Ra 9 . (12) 

Equations (10), (11), and (12) give the travel, velocity, and accelera- 
tion characteristics of the film as a function of the comparable charac- 
teristics of the intermittent mechanism for any position of the pin 
measured in radians from the mid-point of the pull-down period. 

While the equations given above show travel velocity and accelera- 
tion as a function of angular displacement from the mid-point of the 
pull-down period, it is easier to visualize the result if the curves cor- 
responding to the equations are plotted with the abscissa expressed as 
degrees and with zero at the starting point of the pull-down period. 
All of the curves which follow are consistent with the above and the 
number of degrees correspond to the angular displacement of the pin 
shaft or equivalent mechanism which is assumed to operate the inter- 
mittent mechanism once per revolution and which rotates at 1440 
revolutions per minute (24 frames per second) . 

The three curves of Fig. 2A show the characteristics pertaining to 
film driven by the Geneva intermittent mechanism. The upper curve 
indicates film travel, the middle one film velocity, and the lower one 
film acceleration, all as a function of the angular rotation of the pin 
shaft. From the upper curve it is seen that only a very small move- 
ment of the film occurs during the first few and the last few degrees 
of the pull-down. Actually, in the first eight degrees the film is moved 
only 0.005 inch with a like movement occurring during the last eight 
degrees. Inspection of the middle curve shows that the maximum 
velocity of the film is reached at the mid-point of the pull-down period 
and attains a value of approximately 180 inches per second, which is 

Nov. 1947 



about ten times the average speed of the film through the projector. 
The lower curve shows that the film is accelerated for the first half of 
the pull-down. The maximum acceleration or deceleration is approxi- 
mately 60,000 inches per second per second. 

The curves of Fig. 2B show characteristics which are related par- 
ticularly to tension on the film when driven by a Geneva movement. 
The upper curve shows the pull on the film caused by its acceleration 
which is determined by the movement of the intermittent mechanism. 

A B 

FIG. 2. A, Travel, velocity, and acceleration of the film. B, Pull on the 
film caused by acceleration and film-trap drag. 

This pull is determined by multiplying the film-acceleration curve 
by the mass of the film, the determination of which was described 
earlier. In addition to the accelerating pull on the film imposed upon 
it by the intermittent mechanism there is an additional stress due to 
the friction in the film gate. At the mid-point of the pull-down period 
the film is traveling at 180 inches per second and if no means were pro- 
vided to decelerate it, it would tend to continue this speed and 
would try to overshoot the intermittent sprocket. In order that the 
film will never tend to overshoot and will come to rest precisely under 
the control of the intermittent mechanism the film-trap drag must be 
at least equal to the peak inertia force on the film at its peak point of 
deceleration, which is approximately 0.5 pound. The middle curve, 



Vol 49, No. 5 

therefore, shows this additional film-trap drag. Experimental evidence 
confirms this value at least well enough to justify its use in a discussion 
of this kind. The lower curve shows the net pull on the film caused 
by its acceleration by the intermittent mechanism as shown in the 
upper curve and the film-trap drag as shown in the middle curve. 
During the first half of the pull-down the force on the film caused 
by acceleration and film-trap drag act in the same direction and must 
be added. During the last hah 7 of the pull-down they act in opposite 
directions and must be subtracted. As previously indicated, the net 

5,000 FT / SEC 

5,000 FT /SEC 





FIG. 3. Comparison of a constant film- 
acceleration curve with the film-acceleration 
curve of the Geneva. 

pull on the film should never be negative if a steady picture is to be 
achieved. Inspection of the curve shows that the film, is subject to its 
greatest stress during the first half of the pull-down, that this point 
occurs at the point of maximum film acceleration, and that the net 
pull on the film is approximately one pound. 

The thoughts which follow are developed on the assumption that if pull on the film is not allowed to exceed this value of one 
pound, no more damage will be done to the film by other intermittent 
mechanisms than by the Geneva movement and this has been the con- 
trolling factor in the train of thought which follows. 

1. C05/a/ Film Acceleration and Deceleration. Let us now take 
our first step toward the synthesis of an intermittent mechanism 
which will have an operating period of less than the 90-degree period 
of the Geneva movement. Let us see what would be the result if we 
could make an intermittent mechanism, which in association with the 

Nov. 1947 



present type of film trap would give the film constant acceleration 
during the first half of the pull-down and constant deceleration during 
the last half of the pull-down and yet would not impose a net pull of 
more than one pound on the film. 

Let the film be given a constant acceleration of 5000 feet per second 
per second, the same as the peak acceleration of the Geneva. Then 
the acceleration curve of the film as compared with that of the Geneva 
would be as shown in Fig. 3. It can be seen from Fig. 3 that giving the 

-j .5 


* .8 

| 5,000 


t, 5,000 


30 60 


A B 

FIG. 4. A, Travel, velocity, and acceleration of the film. B, Pull 
on the film caused by acceleration and film-trap drag. 

film constant acceleration amounts to chopping off the dotted portion of 
the Geneva acceleration curve and using it to fill in the cross-hatched 
portion of the constant film-acceleration curve. The pull-down angle 
is thus reduced and since the maximum acceleration of the film has not 
been increased there is no increase in the stress on the film. The re- 
duction in the pull-down angle can be computed from the equation 

S = Vi of 
where -5 = ^5 = 1 / a the height of a frame of film 

a = 5000 ft /sec 2 or 60,000 in. /sec 

2s /2X 0.748/2 

^ = \ 60,000 ~ 


9/2 = 360 X 



60 pull-down angle. 



Vol 49, No. 5 

The film travel, velocity, and acceleration curves are shown in Fig. 
4A. The forces on the film due to its acceleration and film-trap drag 
and the resulting net force on the film are shown in Fig. 4B. It can 
be seen that although the pull-down time is reduced, the net pull on 
the film is still the same as that of the Geneva. 

2. Variable Film-Trap Drag. Another possible way to reduce 
the pull-down angle is to vary the film-trap drag. The only purpose 
of the film-trap drag is to decelerate the film during the last half of 


A B 

FIG. 5. A, Travel, velocity, and acceleration of the film. B, Pull 
on the film caused by acceleration and film -trap drag. -. v 

the pull-down. It is not needed during the first half of the pull-down 
when all it does is hinder the acceleration of the film. It would, there- 
fore; be desirable to have film-trap drag only during the last half of 
the pull-down. This might be accomplished by means of an electro- 
magnet which would release the pressure of the film gate during the 
first half of the pull-down. 

If this could be done, then during the first half of the pull-down, the 
pull on the film would be only that due to its acceleration alone ( J /2 
pound). Since we are allowing a maximum pull on the film of one 
pound, the pull on the film due to acceleration alone could be doubled 
and its acceleration increased from 5000 to 10,000 feet per second 

Nov. 1947 



squared. This would permit the time for the first half of the pull-down 
to be reduced to 

0.0026 sec 

2 X 0.0748/2 
10,000 X 12 in. 

., /2 = j|- 6 X 360 = 22'. 

The total pull-down angle would take place in 22 + 30 degrees = 52 
degrees. The film travel, velocity, and acceleration curves would be 
as shown in Fig. 5 A. The forces on the film caused by its acceleration 


22 17 

c " 

- 20,000 

flf'4- -i 

i 1 Lil 

22 17 


A B 

FIG. 6. A, Travel, velocity, and acceleration of the film. B, Pull 
on the film caused by acceleration and film-trap drag. 

and the film-trap drag and the net force on the film are shown in 
Fig. 5B. It can be seen that the net force on the film is still no higher 
than one pound, the maximum pull on the film with the Geneva. 

3. Rapid Deceleration of the Film. A third possible way to reduce 
the pull-down time is to increase the deceleration of the film. During 
the last half of the pull-down, the forces on the film and the film-trap 
drag act in opposite directions so that if they are of equal magnitude 
the net force on the film is zero. If the deceleration of the film were 
increased from 5000 feet per second to 20,000 feet per second squared, 

414 HAYEK Vol 49, No. 5 

the deceleration force of the film would be increased from l / z to 2 
pounds. Then if the film-trap drag were increased from l / 2 to 2 
pounds, the net force on the film would be zero. If it is assumed that 
the deceleration of the film were 20,000 feet per second squared, the 
time for the last half of the pull down could be reduced to 

s = Va at* 

2s 2 X .0748/2 
* = T = 20,000 X i2= 

x 360 = 15 

The time for the full pull-down would thus be 22 + 15 degrees = 37 
degrees. The film travel, velocity, and acceleration curves would be 
as shown in Fig. 6A. The force on the film caused by its acceleration 
and film-trap drag and the net force on the film are shown in Fig. 6B. 
The net pull on the film is still one pound, the same as that of the 


A series of thoughts regarding the operation of an intermittent 
mechanism and its associated film trap have been developed, as pre- 
liminary to experimental design and construction. Of the above 
only the Geneva movement has been used or built in practical com- 
mercial form. The cam of the "Powers" movement is susceptible to 
design so as to give constant acceleration and deceleration and some 
experimental work has been done with a 60-degree intermittent mech- 
anism of this type. 

It should be reaffirmed that this has been written to report on 
thoughts, not things, and to provoke interest in the development and 
use of intermittent mechanisms which will give a shorter pull-down 
time and which will thereby increase the efficiency with which light 
is put upon the screen. 


The author wishes to acknowledge his indebtedness to Mr. G. T. 
Lorance for his presentation of the paper at the SMPE convention 
and for his many suggestions and help in the preparation of this paper. 



Summary. Recorders have been designed to provide the motion picture industry 
with improved lightweight recording machines capable of recording any of the stand- 
ard types of negative or direct positive sound tracks. Separate recorders have been 
designed for use with 35- and 16 -mm film, each recorder being designed for optimum 
performance with the film for which it was designed. Chief features of the recorders 
are improved performance, dependable operation, compactness, minimum weight, 
and accessibility for servicing, combined with attractive styling. The recorders have 
been designed so that the film may be driven from left to right or from right to left 
through the head assembly with equal stability. An automatic film take-up mecha- 
nism of the self -reversing type is provided. 

The PR-32 16-mm recorder and the PR-33 35-mm recorder have 
been designed to provide the motion picture industry improved light- 
weight recording machines capable of recording any of the standard 
types of negative or direct positive sound tracks. Chief features of 
the design are improved performance, dependable operation, com- 
pactness, lightness of weight, and accessibility for servicing. 

The two recorders are identical in design and construction except 
for the difference in the film-handling rollers and sprocket, the take- 
up assembly, the film magazines, and the drive-chain reduction ratio. 

In designing these recorders it was decided that separate recorders 
should be provided for 16- and for 35-mm recording rather than one 
machine should be designed that was intended to be converted for use 
with either type of film. The basis for the decision was the fact that 
each recorder could then be designed without the necessity of making 
compromises which might affect the performance of the entire machine. 

For purposes of economy and simplification of replacement parts, 
the two recorders were designed with as high a percentage of identical 
parts as possible without resorting to undesirable design compromises. 

The recorder (Fig. 1) consists of the following units or subassemblies : 

(A) A base assembly containing the plugs, lamp rheostat, reversing 
switch, and recesses that form the handles for carrying. 

(B) A head assembly containing all of the gearing and film-han- 
dling equipment. 

(C) An optical system. 

(D) A driving motor, single or 3-phase synchronous, alternating- or 
direct-current interlock. 

* Presented Apr. 24, 1947, at the SMPE Convention in Chicago. 
** Radio Corporation of America, Hollywood 38, Calif. 


416 COLLINS Vol 49, No. 5 

(E) A take-up assembly, beltless reversible or belt-drive type. 

(F) A control-panel assembly containing rheostat adjustment, re- 
cording-lamp ammeter, footage counter, and switches for motor, 
lamp, and modulation. 

(G) Covers for the optical system, nibtor compartment, and gear 

FIG. 1. PR-32 16-mm recorder. 

The 35-mm recorder is 23 3 /4 inches long, 9 inches deep, and 2l l / 4 
inches high, with the magazine in place, and weighs approximately 85 
pounds. The magazine is the 1000-foot Bell and Howell type. Other 
type magazines could be used with a suitable magazine spacer. 

,The 16-mm recorder has the same base dimension, but is 18 inches 
high with the 400-foot RCA magazine in place and weighs approxi- 
mately 75 pounds. The RCA 16-mm 400-foot recording magazine is 
of new design using roller-type light traps and pulleys of such con- 
struction that the take-up side may either be belt-driven or driven by 
an engaging pin located in the take-up drive assembly. 

The recorders have been attractively styled and all parts have been 
finished to provide maximum protection against rust and corrosion. 
The recorder base is finished deep umber gray metaluster wrinkle and 


the units above the base are finished light umber gray metaluster 
wrinkle. The film compartment is finished light umber gray enamel. 

Fig. 2 shows the base assembly containing the connection plugs, 
the lamp rheostat, the modulation transformer, wiring, terminal 
board, and built-in carrying recess. The bottom of the base is tapped 
in each corner so that the recorder may be bolted in place if required. 

A continuously variable carbon pile-type rheostat is used to provide 
smooth, stepless lamp control. A two-to-one stepdown control of the 
rheostat is provided to simplify making fine lamp-current adjustments. 

FIG. 2. PR-32/PR-33 base assembly. 

The base casting is made of stabilized magnesium, as are all of the 
structural castings of both recorders. All castings are given a dichro- 
mate treatment before applying the organic finish. This assures 
maximum casting protection and provides excellent adhesion for the 
organic finish. 

The head assembly (Fig. 3) contains all of the film-handling and 
driving equipment except the motor and one silent chain sprocket 
attached to the motor shaft. A step-cut-type light seal is employed 
between the film compartment and its door. A positive-type door 
latch of simple construction and easy operation is located in the bot- 
tom edge of the door. A vertical casting wall divides the film com- 
partment from the drive equipment. 

The film drive of the 35-mm machine consists of a drumshaft and 
flywheel assembly mounted on precision ball bearings, one undamped 
sprung roller assembly, one damped sprung roller assembly, and the 
necessary sprocket-pad roller assemblies. The sprung rollers are 
provided with positive stops to limit their travel, prevent spring dam- 
age, and to assure uniform threading loops. Threading is done with 

418 COLLINS Vol 49, No. 5 

one roller assembly in its normal position as held by the associated 
spring, and with the other roller assembly held against its stop oppos- 
ing the spring action. The sprung rollers are made of anodized alu- 
minum fitted with precision sleeve bearings. The roller arms are ex- 
ceptionally light in weight and are pivoted on precision ball bearings. 

FIG. 3. PR-32 head assembly. 

The pad rollers are held in position against the sprocket by positive 
detents and are held open for threading by spring action. Clearance 
between the pad rollers and the sprockets is controlled by a detent 
plate provided with a simple screw-driver adjustment. 

The recording drum is film-pulled and coupled through its shaft to a 
solid, dynamically balanced flywheel. The sprung rollers are flanged 
to provide the necessary lateral film guiding. 

Threading is very simple and the design is such as to provide the 
same length of film loop each time the machine is threaded. 

The necessary damping is provided by an air-type dashpot coupled 
to the right-hand sprung roller. The design of the dashpot is such 
that no adjustments are required. 

Nov. 1947 



The film drive of the 16-mm recorder is identical to that of the 35- 
mm recorder except for the changes required to handle the narrower 
film and changes in the flywheel inertia and dashpot so as to provide 
optimum motion steadiness at the reduced film speed. 

The film-handling equipment and the gearing have been designed 

FIG. 4. PR-32/PR-33 recorder head, drive side (flywheel 

so that the film may be driven from left to right or from right to left 
through the recorder with equal stability. 

The film-drive mechanism (Fig. 4) consists of a motor mounted on 
the base assembly and coupled to the recorder drive shaft by preci- 
sion chain sprockets and a 3 /i6-mch pitch silent chain, and simplified 
precision helical gearing located in the head assembly. The main 
drive shaft drives the sprocket and the take-up assembly through 



Vol 49, No. 5 

precision-cut right-angle helical gears. A right-angle helical drive from 
the sprocket shaft and a ladder chain are provided to drive the foot- 
age counter. Changes in reduction ratio as required for different fre- 
quencies and different motor speeds are accomplished with both re- 

FIG. 5. PR-32/PR-33 magazine-drive units. 

corders by changing the silent chain sprockets and the silent chain. 
One set of head gearing is used for all applications for both 16- and 
35-mm recorders. 

Fig. 5 shows a beltless reversible-type take-up assembly provided 
for both recorders. This take-up assembly is gear-driven from the 
head and is provided with a thrust-mesh arrangement so that either 
side of the mechanism may be driven. When the recorder sprocket is 
driyen counterclockwise, the right-hand side of the take-up assembly 
is driven and the left side acts as a holdback mechanism. When the 
recorder is reversed and the sprocket is driven clockwise, the take-up 
thrust-gear mechanism reverses and the left side of the take-up be- 
comes the driven side. 

A belt-type take-up is provided as optional equipment and is de- 
signed to mount interchangeably with the beltless take-up. With this 
take-up it is necessary to move the belt manually to the proper maga- 
zine pulley when the recorder direction is reversed. 


The optical-system compartment (Fig, 6) is provided with a hinged 
door through which all normal operating adjustments are made to 
the optical system except for galvanometer tilting, which is done with- 
out opening the door, by an adjusting screw accessible from the left 
end of the optical-system cover. The complete optical-system housing 
is removable by loosening two knurled thumbscrews located in the 

FIG. 6. PR-32/PR-33 optical-system compartment 
(housing removed). 

compartment. The optical-system door is equipped with an opening 
for observing the visual monitor. 

The housing (Fig. 7) covering the back of the control panel and the 
motor compartment is removable by loosening two screws located in 
the end of the housing. This cover is provided with a recess for a 
flush-type handwheel mounted on the motor shaft. When the motor- 



Vol 49, No. 5 

compartment cover has been removed, the motor, drive chain and 
sprockets, footage counter and drive, rheostat control, terminal 
board, and all switches are readily accessible. 

Both recorders may be provided with single-phase 115-volt syn- 
chronous motors, 230-volt, 3-phase synchronous motors, alternating- 
current interlock (Selsyn) motor, or multiduty (direct-current inter- 

FIG. 7. 

PR-32/PR-33 motor and control compartment 
(housing removed). 

lock) motors; however, the 16-mm recorder normally will be supplied 
with a single-phase, 115-volt synchronous motor. In order to keep 
vibration and noise to a minimum, all motors used are statically and 
dynamically balanced and are resilience-mounted to the recorder 
base. The motor is provided with sufficient lateral adjustment to 
assure proper tension of the drive chain and to compensate for any 
possible increase in length of chain after considerable service. 


The cover plate over the gear compartment is easily removed for 
inspection or service. When the cover has been removed, the fly- 
wheel, dashpot assembly, and all gearing are readily accessible. 

The optical systems (Fig. 8), which have been especially designed 
for these recorders, provide rear-projection visual monitoring of an 
improved type. The visual monitor adjustments have been greatly 
simplified and the monitoring screen, as well all mechanical and op- 
tical parts of the monitor, is mounted to the optical-system casting. 
The monitoring screen and mirror assembly are removed as a unit 
from the optical system when focusing the system or when it is de- 

FIG. 8. PR-32/PR-33 optical system. 

sired to remove or insert the ultraviolet filter. Since the monitoring 
screen and mirror assembly is doweled in place, it can be removed and 
reassembled on the system without disturbing any adjustments. The 
optical system used for 35-mm recording and the system used for 16- 
mm recording are identical except for the dimensions of the recording 
slits and apertures. The optical system has been designed to provide 
a maximum of hand room while compactness is retained. The re- 
cording lamp is a prefocused base, 10.5-volt, 7.8-ampere, curved, 
coiled-filament lamp. An improved lamp socket, providing the nec- 
essary vernier adjustments, is used, as is also an improved modulator, 

424 COLLINS Vol 49, No. 5 

exceptionally sturdy and reliable. The area-type optical system is 
standard with both recorders; normally it is provided to produce a 
biased- type standard track. Both optical systems provide for stand- 
ard negative recording, direct positive recording, or reversal-type 
track without requiring any modification or adjustments to the sys- 
tem, except for tilting the modulator. A focusing microscope is pro- 
vided for both optical systems. The microscope is inserted in the re- 
cording drum and provides a very useful operating and inspection 
tool. The optical systems are extremely flexible in application anjl 
may be supplied for any of the listed types of recording : 

(A) Standard area (D) Class AB push-pull 

(B) Class B push-pull (E) Double-width recording 

(C) Class A push-pull (35-mm system only) 

(F) Direct positive 
If desired, a variable-intensity-type optical system can be provided. 


Studies of film motion made by flutter measurements on both the 
16- and 35-mm model recorders indicate that the performance of the 
machines will compare very favorably with lightweight recorders 
available to date. The total flutter for all frequency bands is approxi- 
mately 0.05 per cent. The sprocket-hole frequency disturbances 
have been reduced to a negligible value. The low-frequency disturb- 
ances have been kept to a minimum by careful designing of the filter 
components and film-drive equipment. 


In the design of these recorders every consideration has been given 
to the problems of operation, maintenance, compactness, and sturdi- 
ness. All parts have been amply designed so as to insure reliability 
and at the same time the weight and bulk have been kept to a mini- 
mum. Adjustments have been eliminated wherever possible and the 
necessary adjustments, which have been retained, are simple to make 
and easily accessible. 

The same high degree of accuracy and precision is used in making 
the parts of these recorders as is employed in the manufacture of parts 
for the Type PR-31 De Luxe Film Recorder. Selective assembly is 
not tolerated and all components are held to close tolerances to assure 
interchangeability of parts and optimum performance^ 

It is our belief that the PR-32 16-mm recorder and the PR-33 
35-mm recorder will completely satisfy the requirements for high- 
quality, lightweight recording machines. 





Summary. It has been necessary in the past to control the color balance and ex- 
posure of 16-mm duplicating printers by making actual test prints at frequent inter- 
vals. A large amount of footage could be risked during the time required for process- 
ing the test prints. A method is described by which it is possible to check the balance 
of the printer instantly by means of tricolor readings with a photronic cell. 

The printing of 16-mm Kodachrome duplicates on available mo- 
tion picture printing equipment involves many problems which make 
this pperation considerably more complex than the printing of black- 
and-white film on the same equipment. In black-and-white 
work, density alone has to be controlled in printing, while 
in making Kodachrome duplicates, both density and color balance 
must be controlled. 

This control must be very accurate, since small changes in color 
are more noticeable than similar small changes in black-and-white. 
Density control in black-and-white printing can be accomplished by 
aperture changes or, more conveniently, by changes in lamp voltage. 
However, in Kodachrome duplicating, density must be controlled by 
aperture only, or by the use of neutral-density filters, since voltage 
changes will affect color balance. 

Color balance in this work is controlled by the use of gelatin color- 
compensating filters. These are cyan, magenta, and yellow filters 
available in several small color-density steps. The color-balancing of 
the printer is accomplished by adding or subtracting combinations 
of these filters until the desired balance is obtained on the print. 

In setting up a printer for black-and-white printing it is only nec- 
essary to make a range of exposures on the stock to be used and to 
pick the proper exposure by density measurements from the resulting 
prints. The setting up of a printer for Kodachrome duplicating, how- 
ever, is more difficult. The lamp voltage first must be adjusted to 

* Presented Apr. 22, 1947, at the SMPE Convention in Chicago. 
** Eastman Kodak Company, Kodak Park Works, Cine Kodak Processing 
Department, Rochester, N. Y. 


426 AEX Vol 49, No. 5 

give a color temperature and lamp brightness which will make a 
satisfactory balance possible. The proper exposure range can then be 
determined by a series of flash tests or actual prints, the actual expo- 
sure changes being made by aperture or by additions of neutral-den- 
sity filters, so as not to change the color temperature. After this has 
been done it is necessary to start out with some basic filter combina- 
tion and to make a series of tests, adding or subtracting color-com- 
pensating filters, until a neutral balance is obtained. This usually en- 
tails printing a number of actual tests on the Kodachrome duplicating 
stock. Specific instructions for the procedure are contained in the 
manual "Instructions for Making 16-Mm Kodachrome Duplicates 
on Kodachrome Duplicating Film, Type 5265", which may be ob- 
tained from the Motion Picture Film Department of the Eastman 
Kodak Company. If several printers are to be used, it is at present 
necessary to go through the procedure described, to balance each one. 

After a printer has been satisfactorily balanced to make 16-mm 
Kodachrome duplicates, it must be checked periodically to ensure 
keeping it in balance. This requires a constant check on the quality 
of the production work printed, and also the making of test prints 
for the specific purpose of a critical check on color balance and density. 

This method of checking color balance and density on Kodachrome 
duplicating printers is necessarily time-consuming. The problem is 
especially serious when the printer is located at some distance from 
the processing laboratory. Since it is generally necessary to keep the 
printer operating continuously, the delay involved in the transporta- 
tion of the work to and from processing may necessitate the risk of a 
considerable amount of footage before the results of tests or current 
production can be seen. If the quality of the finished work, or the 
tests indicate that the printer is off-balance, further filter changes and 
tests must be carried out and additional delay is encountered. 

Because of these problems there has long been a need for some sim- 
ple "and rapid method to check Kodachrome duplicating printers. 
An instrument has been devised recently which will provide an accu- 
rate and almost instantaneous test of both color balance and exposure 
on such printing equipment. 

The instrument consists essentially of a photocell and a sensitive 
galvanometer arranged so that light-readings can be made at the 
printer aperture through red, green, and blue filters. These read- 
ings are an indication of both color quality and exposure. Such an 
instrument is illustrated in Fig. 1. 


As pictured, the instrument has been constructed for use with a 16- 
mm Depue continuous printer. Variations, however, could be made 
for use with almost any type of 16-mm printing equipment. The 
photocell housing in this case has been shaped to fit snugly over the 
gate and aperture of the Depue printer so that stray light is elimi- 
nated. The aperture of the photocell housing has been made to ex- 
actly the same size as the printer aperture and the housing is designed 
to fit in place in only one position, so that the aperture of the printer 
and that of the photocell will coincide. 

In order to minimize the effects of cell fatigue and consequent error, 
it has been found necessary to distribute the light striking the photo- 
cell evenly over the entire cell surface. In this particular case, the 

FIG. 1. Photocell head and galvanometer. 

limited amount of space available at the gate of the Depue printer 
makes this difficult, but the method by which it is accomplished is 
shown in the drawing of the photocell housing, Fig. 2. 

The light entering the aperture A is gathered up by a Lucite rod B 
and is conducted to the cell C. The end of the lucite rod D is rounded 
and the surface has been ground so that it projects the light evenly 
over the entire surface of the cell. A sliding carrier R holding the 
red, blue, and green filters is located between the rounded end of the 
rod and the cell. In this way the three-color readings can be made 
simply by sliding the carrier so as to place first the blue, then the 
green, and finally the red filter in the beam. 

The photocell used in this case is an Electrocell No. 2, regular type, 
2 inches in diameter. The sensitivity curve for this cell is very simi- 
lar, however, to a Weston Photronic cell, Type III, and it is planned 
eventually to substitute the Type III for the present one. 



Vol 49, No. 5 

The cell is connected to the galvanometer by a shielded lead to 
prevent errors from outside electrical disturbances. The galvanom- 
eter* used is equipped with an adjustment for zero setting and has 
a scale reading from to 80. A sensitive galvanometer of this type is 
necessary in order to get sufficient accuracy because of the small light 
intensities measured, and consequent small currents generated by the 
cell. This particular combination of Electrocell and galvanometer 
has proved to be extremely sensitive to light of low intensity such as 
is encountered in making printer measurements. Although the 
G. M. galvanometer is a very sensitive instrument, -it has proved 
to be satisfactorily resistant to disturbance and shock resulting 
from reasonably careful handling. 

FIG. 2. Enlarged view of photocell head. 

The tricolor niters chosen for use with this instrument are the 
Wratten No. 47 blue, No. 53 green, and No. 29 red. These filters in 
combination with the Electrocell give sensitivity peaks in the blue, 
green, and red portions of the spectrum which correspond to the criti- 
cal sensitivities of the Kodachrome duplicating film. Readings made 
at the printer aperture, therefore, indicate the relative intensities of 
the blue, green, and red components of the printer light. Proper ad- 
justment of these intensities by use of the color-compensating filters 
in the printer will result in a neutral balance on the Kodachrome film. 
The instrument is sensitive enough to record a substantial change in 
reading when the most dilute color-compensating filters are added or 
subtracted from the printer. It will also record a one-sixth-stop 
change in exposure. 

* G. M. mirror type, Serial No. 9640, Catalog No. 11026, G. M. Laboratories, 
Inc., Chicago, 111. 



The instrument in use on a printer is shown in Fig. 3. Before lo- 
cating the photocell housing on the printer, the galvanometer should 
be adjusted to zero. The printer lamp previously should have been 
turned on and allowed to burn for about two minutes. This permits 
the lamp to come to equilibrium. The printer aperture should be 
set at the normal opening for printing from normally exposed origi- 
nals. The aperture of the photocell housing is then located over the 

FIG. 3. Photocell head on printer. 

aperture of the printer and the cell is clamped securely in place. The 
location of the photocell housing on the printer, as shown, is facili- 
tated by a pair of stops on the housing which ensure accurate align- 
ment of the two apertures. When the cell is in place, the blue filter is 
placed in the beam and the galvanometer reading is taken. The 
green filter reading is taken next, and then the red. 

In order to use the instrument properly, one printer must first be 
balanced to give a neutral color balance and correct density on the 
Kodachrome duplicating film by photographic tests, as has been 

430 AEX Vol 49, No. 5 

previously described. Once this has been done, however, and tricolor 
readings taken with the instrument, these readings may be used as a 
standard. Additional printers of the same type may then be balanced 
simply by comparing the new printer with the instrument and by 
bringing the readings to match the standard. 

When the printers have been balanced and are in production, 
they may be checked periodically with the instrument, thereby giving 
an instant check on the quality of the production work. If this check 
indicates that a printer is out of balance, it may immediately be 
brought back to standard by making the necessary printer changes. 
To facilitate this it has been found desirable to compile tables showing 
the changes in tricolor readings caused by the addition or subtraction 
of individual color-compensating filters to the printer. Similar tables 
have been made for exposure changes. In this way, when an off- 
standard reading is obtained, the proper corrections can immediately 
be calculated and applied. 

Individual emulsion numbers of Kodachrome Duplicating Film, Type 
5265, may require slightly different balances. .When a new emulsion 
number is to be used, one printer will have to be balanced by a pho- 
tographic test and a new standard reading for this emulsion deter- 
mined. This, of course, is simplified by the information given on the 
film carton concerning the filter change required for the emulsion 
batch. When the new standard reading is obtained the other printers 
may be changed over to the new emulsion by use of the instrument. 

Under actual production conditions the instrument has proved to 
be an extremely valuable aid in the control of 16-mm Kodachrome 
duplicating printers. In several cases when used for setting up new 
printers it has given a satisfactory balance on the first test. Compared 
to the numerous tests required to balance a printer by strictly pho- 
tographic methods this represents a considerable saving in time and 
effort. The instrument has been used for some time for the control 
of several 16-mm Depue .printers engaged in continuous production of 
Kodachrome duplicating work. In this operation it has provided an 
accurate, instant check on color balance and exposure, and has 
eliminated the use of daily photographic printer tests. Thus, it has 
saved time, lowered waste, and reduced cost. 



Summary This paper describes three types of loudspeaker systems which have 
been designed for 16-mm sound on film reproduction. One system uses a dia-cone 
speaker mounted in a portable leatherette carrying case. A 12-inch speaker is used 
and the acoustic radiation is obtained from two diaphragms attached to a single voice 
coil. The outer diaphragm is of the molded-cone type and the inner is a domed dural 
type of diaphragm which is attached at its edges to the 3-inch voice coil. A second 
system is mounted in a portable cabinet which has 3.2 cubic feet and is resonated for 
maximum response down to 85 cycles. The speaker is of the 15-inch type and of 
similar construction to the 12-inch with the addition of a 6-cell multicellular horn 
mounted directly in front of the 3-inch dural diaphragm. The third system consists 
of a duplex speaker and is intended for permanent and semipermanent installations. 

The widespread application of 16-mm sound on film in the enter- 
tainment, educational, and industrial fields has created demands for 
higher quality. In the past, quality has been limited by various fac- 
tors such as flutter content, size of the image in relation to the slit, 
inadequate power-amplifier output, and inefficient limited-range 

This paper describes three types of loudspeakers which are designed 
to utilize more fully the new improvements in 16-mm projectors 
which shortly will be made available. 

The principal limitations in 16-mm loudspeakers are caused by the 
low efficiency, lack of uniform distribution over a wide angle, inade- 
quate frequency range, and high distortion at operating levels. Three 
different types of units are available varying in size and weight for 
portable and semiportable installations, which minimize these defi- 

In order to provide a lighter-weight and lower-cost unit which re- 
tains most of the good features of the two-way loudspeaker, the de- 
sign now known as the dia-cone was developed. The name dia-cone 
is derived from "diaphragm" and "cone" and applies to a loud- 
speaker having both a high-frequency diaphragm and a low-frequency 

* Presented Apr. 22, 1947, at the SMPE Convention in Chicago. 
** Altec Lansing Corporation, Hollywood, Calif. 




Vol 49, No. 5 

cone driven through a mechanical network by a single large voice 
coil. The combination thus gives many of the advantages of a true 
two-way loudspeaker without the accompanying high costs of double 
magnets, double voice coils, crossover networks, and the additional 
costs necessitated by a complicated mechanical construction. 

The Model 603 multicell 
dia-cone loudspeaker (Fig. 1) 
employs a 3-inch voice coil 
having an inner diaphragm 
made of dural, and an outer 
cone which is made of seam- 
less molded felted paper. At 
frequencies above 2000 cycles 
the mass of the outside cone is 
very large and, as a conse- 
quence, its ability to radiate 
sound uniformly decreases as 
the frequency increases. At- 
tached to the voice coil di- 
rectly is a domed metal dia- 
phragm 3 inches in diameter. 
This metal diaphragm has a 
high stiffness-mass ratio and 
high sound-transmission speed. 
It is able to operate as a 
piston even though the out- 
side cone fails to provide the 
proper movement. The voice 

coil and inner diaphragm vibrate independently of the outer dia- 
phragm at the higher frequencies because of the compliance in the 
cone immediately outside the area and adjacent to the voice coil. 
The. vibrating area of the metal dome is small in comparison with 
the wavelengths of the frequencies being radiated, and for this 
reason the angle of distribution is considerably widened over the 
single cone. The amplitude of the high-frequency diaphragm for 
uniform radiation of acoustic power decreases with an increase of fre- 

For this reason, considerable acoustic power can be radiated at the 
higher frequencies from the 3 inch diaphragm with a comparatively 
small excursion. At low frequencies, this center portion vibrates in 

FIG. 1. Side view of Model 603 
multicell dia-cone loudspeaker. 


phase with the outer diaphragm and provides the maximum possible 
vibrating area. In order to enhance further the distribution pattern 
over the high-frequency range, a cast-bakelite 6-cell horn is mounted 
directly in front of the metal dome, as shown in Fig. 2. A clearance of 
about 150 mils is provided so that at its maximum rated power there 
is no danger of the metal dome's striking the throat of the horn. 
The multicellular horn is held in position by means of two studs which 

FIG. 2. Front view of Model 603 multicell 
dia-cone loudspeaker. 

are threaded into the top plate surrounding the voice coil and pole- 
piece structure. Clearance holes are provided in the cone for these 
studs. Acoustic-performance tests indicate that the multicellular 
horn greatly improves the radiation pattern of the loudspeaker and 
provides sufficient loading to reduce the irregularities in response. 

An edge wise- wound aluminum-ribbon voice coil is used. The use 
of edgewise-wound ribbon improves the space factor over that of 
round wire, and since more conductor material can be placed in the 
air gap, the efficiency is raised and the operating temperature, with 
higher power, correspondingly decreased. Since the 3-inch voice-coil 
diameter is considerably larger than the voice coil normally used on 



Vol 49, No. 5 

loudspeakers intended for this service, it has correspondingly in- 
creased ability to handle higher power without undue temperature 
rise, and as a result, the efficiency is little affected by changes in 

The outer rim of the cone is cemented to the frame. The central 
spider assembly is attached to the cone outside of the voice coil and is 
of the accordion type, so as to permit large low-frequency excursions 

with low distortion. The res- 
onance of the cone and voice- 
coil assembly is approximately 
55 cycles in free air. 

An Alnico V permanent 
magnet is provided for the 
field excitation, and the total 
energy available with this 
magnet is greater than that 
previously supplied in the 
field-coil-type loudspeakers. 
The magnet itself is of the 
center-core type. The soft 
magnetic material, forming 
the path between the pole 
pieces, is amply designed so 
that the flux is conducted 
through the outside walls and 
up to the air gap where the 
voice coil is mounted, with 
little loss. The external leak- 
age loss is extremely low. Ad- 
ditional benefits from these features of the design are increased effi- 
ciencies owing to lower magnetic losses, and the fact that it is 
possible to handle this unit without endangering wrist watches 
or other devices which may be susceptible to damage from magnetic 

The Model 600 dia-cone loudspeaker is similar to the Model 603 
just described. (See Fig. 3.) It is mounted on a 12-inch frame, and 
the vibrating system has an area of 67 square inches as compared to 
the 123 square inches for the 603 unit. In order to reduce the weight 
and cost for portable application, the multicellular horn is not used. 
(See Fig. 4.) 

FIG. 3. Side view of Model 600 
dia-cone loudspeaker. 


This represents some compromise in response and distribution 
compared to the 603 unit. Both the 603 and 600 speakers have an 
impedance of approximately 10 ohms at 1000 cycles, and their effi- 
ciency is such that they will deliver 89 decibels (reference 0.0002 dyne 
per square centimeter) on its normal axis at 5 feet with an input of 
0.1 watt. 

The Model 604 duplex loudspeaker has been described in a previous 
meeting of the Society. Briefly, it is a two-way speaker using indivi- 

4. Front view of Model 600 dia-cone loudspeaker. 

dual diaphragms, voice coils, and magnets for each unit. An elec- 
trical dividing network is used and the frequency crossover is approxi- 
mately 2000 cycles. 

This loudspeaker is 3 decibels more efficient than either the 600 or 603 
speakers in the range from 100 to 2000 cycles, and considerably more 
efficient beyond this range. (See Fig. 5.) It is being used in fixed and 
semiportable installations where the highest quality is desired. Club 
cars and dining cars on railroad systems are using this loudspeaker in 
connection with their 16-mm-film entertainment equipment. Other 
applications include schools, clubs, and private homes. 



Vol 49, No. 5 

When using 600 and 603 loudspeakers with amplifiers having nega- 
tive feedback which includes the output stage, the maximum true 
bass response can be obtained when the internal output impedance of 
the amplifier is approximately 10 ohms. It is not alone sufficient that 
the amplifier be rated for a 10-ohm load, since the use of very large 

amounts of feedback may 
produce output imped- 
ances very much lower 
than the rated impedance 
of the amplifier. 

An amplifier-output 
impedance, several times 
lower than the speaker 
impedance, should be 
used only in connection 
with loudspeaker cabi- 
nets which are of im- 
proper design, producing 

Cabinets of the turned- 
port type are recom- 
mended where the maxi- 
mum bass response is 
FIG. 5. Side view of Model 604 duplex 

loudspeaker. required consistent with 

limited space. 

The frequencies indicated on Fig. 6 afe recommended points of 
resonance for the indicated volume of the cabinet. 

If, for any reason, it is desirable to change the frequency of reso- 
nance of the port, the following procedure can be used : 

1. Select the frequency where the port is to resonate or provide 
maximum response. 

2. By means of an audio oscillator and amplifier of correct output 
impedance, provide approximately 1 watt at the voice-coil terminals 
at the selected frequency. 

3. Place a volume indicator, vacuum-tube voltmeter, or other 
measuring device across the terminal of the loudspeaker. 

4. Adjust the area of the port until a minimum deflection is ob- 
tained on the meter. This area then provides the maximum acoustic 
response possible for the selected size at the measured frequency. 

It is desirable to mount the speaker as high in the cabinet as 

Nov. 1947 



possible. This height is necessary so that the direct radiation from 
the loudspeaker will not be obstructed by furniture, and so that re- 
flections from the floor will be minimized. 

Fig. 7 shows the 612 utility cabinet which has a volume of approxi- 
mately 6 cubic feet, and the port is tuned to 60 cycles. It is also lined 
with fiber-glass panels. 

Fig. 8 shows the 614 portable cabinet which has 3.2 cubic feet, and 
is resonated for maximum response down to 85 cycles. This cabinet 
is intended for portable public-address and 16-mm service. 


CU. FT. 

CD. IN. 

8650 10350 





FIG. 6. 

Chart showing area of port versus cabinet volume. 

In order to demonstrate these loudspeakers, especially prepared 
16-mm prints were made and reproduced on a projector which had 
been designed around the Joint Army and Navy Specifications, pre- 
pared by the ASA-Z52 War Photography Group. The over-all re- 
sponse of the entire system, as measured with the Z22.44-1946 16- 
mm test film, is uniform within 1 decibel in the low-frequency region 
down to 60 cycles. In the high-frequency region, the response was 
uniform up to 4000 cycles and 2 at 5000 cycles and 8 at 7000 cy- 
cles. The total flutter, when measured according to the latest stand- 
ards of rating flutter, is not greater than 0.1 per cent. Additional 
information on the frequency test film used and flutter measurement 



standards can be obtained from 
the office of the Society of Motion 
Picture Engineers. 

One sample of film demonstrat- 
ing solo voice and piano was 
originally recorded on 200-mil 
push-pull variable density, and 
then re-recorded to a 16-mm 
negative from which the final 
print was made. The response 
of the re-recording system was 
set so as to be approximately the 
same as that of 35-mm technique. 
The second selection was a 20- 
FIG. 7. 612 utility cabinet. minute short subject. This film 

was re-recorded from a 35-mm variable-area track on to a 16-mm 
negative. Some manual compression was used to reduce the volume 
range at the time of re-recording. The processing on the print was 
adjusted so as to give a minimum of 30 decibels cross modulation. 

The demonstration was given 
in the Esquire Theatre, which 
has a seating capacity of approxi- 
mately 1500. The amplifier sys- 
tem used had an installed capa- 
city of 20 watts. The efficiency 
of the loudspeaker is such that 
approximately 5 watts of electri- 
cal power were required to pro- 
duce the required acoustic energy 
considered adequate on the basis 
of 35-mm presentation of musi- 
cals. The demonstration used a 
Model 603 loudspeaker mounted 
in a. 6 14 cabinet, having an enclo- 
sure of approximately 3.2 cubic 
.. FIG. 8. 614 portable cabinet, 


The results of this demonstration indicate that when all of the ele- 
ments of the system such as recording equipment, projector, and 
loudspeaker meet certain minimum standards, 16-mm quality can be 
competitive with 35-mm sound systems. 



Summary. Source, size, lens aperture, steadiness, film resolution, and similar 
factors govern the present-day and anticipated future limits of 8-mm motion pictures 
as to dynamic resolution, illumination, and possible sound on film. 

In order to evaluate the subject to screen performance of motion picture systems, a 
test method was employed which indicated the total resolution of the camera, film, and 
projector. In order to record resolutions equivalent to those seen by eye, it was found 
necessary to add the images of several frames because of persistence of vision. The 
same conditions appear to apply to television. 


In the absence of basic papers dealing with the problems of 8-mm 
equipment and performance, there appears to be a need for the re- 
examination of the basic precepts in view of recent developments and 
trends. One of the peculiarities of the field seems to be the apparent 
disregard for the need of engineering knowledge in a class of equip- 
ment operating at magnifications equivalent to most optical micro- 
scopes. The probable reason for the lack of engineering study of 8- 
mm equipment is the apparently prevalent concept that such equip- 
ment is largely a problem requiring only mechanical engineering to 
the exclusion of the optical and electronic knowledge. An integration 
of all three is required as in motion picture engineering practice with 
other size films. 

A crude survey of equipment costs shows that there is little differ- 
ence in the cost of 8-mm apparatus as compared to 16-mm apparatus 
of similar manufacturing quality; in fact, in some instances the 
mechanisms are somewhat interchangeable. The fundamental de- 
sign character then rests largely with the film size itself. The econ- 
omy is due to more than a small image, because of the use of 16-mm 
processing equipment, handling 8 mm as dual 8 mm prior to splitting. 
It is possible to conceive a considerable number of arguments for 
films of other dimensions which will allow greater economy by more 
efficient use of the emulsion area, such as 9.5 mm which allows a frame 
image very nearly as large as 16 mm, or a possible 4-mm size which by 
the saving of area by the use of a notch-and-ratchet pull-down may 

* Presented Apr. 25, 1947, at the SMPE Convention in Chicago. 
** Armour Research Foundation of Illinois Institute of Technology, Chicago 
16, 111. 


440 LEWIS Vol 49, No. 5 

have a frame size the same as 8 mm. However, the fact remains that, 
unless considerable economy over the present 8-mm size is possible, 
the quantity of production will not be sufficient to guarantee low 
camera and projector sales prices. If, on the other hand, the addition 
of sound to 8 mm is contemplated, the reduction of the presently un- 
used emulsion area must be considered differently. It is obvious that 
8-mm film was conceived for silent-picture use, or the frame would 
have been rotated 90 degrees to allow faster film velocity for the same 
picture-area utilization. 

The fundamental characteristic of 8-mm film is then largely due to 
the use of a small frame which allows economy of film, though at the 
same time using available 16-mm processing stations. As the frame 
size is close to one fourth that of 16-mm film, one would expect a 
screen sharpness one quarter that of 16 mm with other factors the 
same. From the strictly interpreted viewpoint, the comparison of 
subject to screen performance resolution is the final criterion. There 
is, however, another possibility which permits a definitely different 
viewpoint; namely, whether the system will produce a satisfactorily 
sharp image over a screen of 8 by 10 inches, for example. When con- 
sidered from the standpoint of a screen of limited dimension, reso- 
lution above the acceptable point becomes superfluous. If resolution 
equivalent to newspaper quality is considered acceptable, then ap- 
proximately 500 horizontal lines are required. The design of a pro- 
jector under such considerations then becomes a cross between an 
editor and a "juke box", similar to a desk-model television receiver. 
That 8 mm is to television what the phonograph is to amplitude - 
modulation radio seems logical. 

In view of the above discussion, the following is a survey of the fac- 
tors which influence the performance of 8-mm equipment, in terms 
of design quantities and fundamental equations. 


Film for 8-mm cameras may be purchased in three types of load- 
ings : single-eight, double-eight, and double-eight magazine. Eight- 
millimeter projectors use single-width 8-mm film universally, with 
loading being by means of manual threading from standard reels. 
Neither magazine loading nor automatic threading for 8-mm pro- 
jectors has appeared on the market as yet. 

Though one 8-mm frame requires one quarter the area of a 16-mm 
frame, the current cost per foot of 8-mm film is not of that ratio, but 


on the projection time basis more nearly two to one. Thus on the 
number of lines resolved per dollar basis, the 8-mm user now gets a 
value of about one half that of the 16-mm user. 

From the standpoint of comparative performance, an 8-mm image 
cannot win, yet for an acceptable image for home use, the frame is 
apparently adequate. If a film resolves 60 lines per millimeter, 180 
lines per frame will be resolved for television scanning or 240 when 
scanned the other way at right angles. Microfilm emulsions resolve 
120 lines per millimeter giving 360 by 480 lines per image. Emul- 
sions resolving as high as 500 lines per millimeter have been made for 
spectroscopic plates (Eastman Kodak Company, Type 548), which if 
coated on 8-mm stock would give 1500 by 2000 line images. The 
use of an emulsion as slow as the last would not be practical for 
camera, though by reduction printing from 35 mm, full realization of 
its potential could be had. 

If the perception of movement on the screen by the eye is assumed 
to be at a limit at 16 frames per second, and the perception of flicker 
or intensity modulation disappears at three times that, or 48 cycles 
per second, we have a basis for further investigations of image con- 
ditions. If these conditions are properly met, we will then have three 
interruptions per frame at silent speed so that in two sixteenths of a 
second, two frames appeared three times each for a total of six. Also, 
if we choose at random an interval of one sixteenth of a second, we 
may have an interval containing one showing of the first frame and 
two of the second and so on. Consequently, the eye actually sees a 
mixture of two or more frames. If two frames are seen, it is im- 
probable that persistence of vision would cause an improvement in 
resolution equal to more than the number of frames taken as a factor ; 
in this case, twice. For 60-line-per-millimeter film, this resolution 
would result in a picture of 360 by 480 lines and so on. For the same 
film at sound speed, three frames may be seen enough to be visually 
integrated to resolve not more than a 540 by 720 line image. At tele- 
vision speed, 30 frames per second, a picture of fully twice and prob- 
ably nearer three times resolution is possible as a limit. Needless to 
say, if the mechanism does not place each frame in exactly the same 
place, image deterioration will result, which result is due also to per- 
sistence of vision. 

From the theory of persistence of vision with regard to resolution, 
we see that a finer line may be shown from the same emulsion when 
used as a motion picture rather than a still picture, assuming all other 

442 LEWIS Vol 49, No. 5 

factors the same. Therefore, inasmuch as single frames of 8-mm film 
will appear when seen as stills, half as sharp as the motion picture, 
there will in all probability be little demand for stop-on-film construc- 
tions popular for 9.5-mm film in Europe. 



Lines per Millimeter Horizontal Lines per Image Vertical Lines per Image 

60 540 720 

120 1080 1440 

500 4500 6000 


As films resolving 60 lines per millimeter are commonly in use to- 
day, the persistence of vision effect alone would seem to indicate that 
lenses resolving at least 120 to 180 lines per millimeter are required for 
full use of the resolution potential by both the camera and the pro 

For microscope use, lenses resolving 300 lines per millimeter over 
the entire frame are currently available with exposure-control dia- 
phragms. The similarity between 8-mm lens and the microscope 
objective does not end here, for the element sizes,. centering problems, 
and similar problems are very nearly the same. However, the 8-mm 
lens at present is unnecessarily restricted as to back clearance and out- 
side diameter. As the fields are nearly the same and the magni- 
fications required of the same order, it appears logical that the numeri- 
cal aperture of the objective will determine its resolving power when 
abberrations are corrected. Therefore, the faster the lens, or the 
larger the numerical aperture, the higher the resolving power. It 
seems in order at this point to predict that the present commercial 
limit of relative aperture (//1. 4) will soon be opened up in favor of 
sharper pictures as well as more favorable exposures. 

This will not be possible if the present back clearances must remain 
so great that the microscope design types of lenses are ruled out. 
Lift-out turrets and back-opening gates offer an obvious answer. 
Likewise, high relative aperture designs and wide-angle designs are 
very often made useless by the poor mechanical clearances allowed 
for the lens mounts. All too often the lenses of a turret are so close 


together that one shows the other in the picture, if the other can be 
mounted at all. The economies of small gears often cause projector 
designers to place the sprockets so close together that the outside 
diameter of the lens is restricted to the diameter of the particular 
Petzval objective the designer intended using. This often makes the 
use of an anastigmat or other type of objective impossible. This type 
of error is all too characteristic of the mind that knows nothing out- 
side of mechanical engineering and, therefore, never does a complete 
job of motion picture engineering. 

Eight-millimeter cameras are particularly fortunate in one respect; 
because of the small picture size, the focal lengths of the lenses are 
short and thereby a great depth of focus is acquired. This enables 
practical use of relative apertures impossible in longer focal lengths. 
There is another side to the same story, however; a great shift in the 
object distance results in a small shift in the image distance; a small 
shift of the same order of the lens away from the film results in a great 
shift of the median focus in object space. As a result of this effect, 
8-mm cameras are on the market which, according to the combination 
of magazines, lenses, turret, or whatever else is involved, the "in- 
finity" focus may be anywhere from 6 feet to beyond infinity (if such 
a term exists). 

Others have demonstrated quite adequately that with known tung- 
sten sources little improvement can be expected of present-day 
illumination systems for projectors, because of the restricting effects 
of filament size, bulb diameter, frame size, and lens aperture. In 
short, only by the use of objectives of higher relative aperture and con- 
densers to match, can more lumens be supplied to the screen from 
tungsten lamps. 

Arc sources as yet have not been demonstrated to be commercially 
attractive in the 8-mm field. On the basis of area of the aperture, 
using a 16-mm arc delivering 1000 lumens to the screen, it is probable 
that at least 250 lumens would reach the screen for 8-mm use. The 
zirconium concentrated arc shows promise but no data are as yet 
published concerning the lumen output for 8-mm projection with it. 
Until better screen illumination is possible, it is therefore again neces- 
sary to point out that 8-mm screen images are best kept small. 

In the struggle to extract all of the lumens possible out of the pro- 
jector, design of the shutter is often special in an effort to take advan- 
tage of the flicker frequency or other characteristics. The usual 
result of such special designs, which depend on interrupter blades 

444 LEWIS Vol 49, No. 5 

smaller than the pull-down blade, is to produce nicker by visual dis- 
crimination between long and short interruptions. It is common in 
projectors to use pull-down times equal to two-skip or more move- 
ments in order to obtain greater light transmission without travel 
ghost. The same skip design as yet is not used in cameras wherein 
the pull-down usually occupies nearly 50 per cent of the cycle (Vso of 
a second for silent speed), where an exposure of l / m second or longer 
is possible. 

It is all too recurrent a practice that the lamphouse of a projector 
is so designed that the operator may blind himself by looking at a 
brightness of 1000 f6ot-lamberts or more so that the image on the 
screen is no longer visible to him. If he is fortunate enough to escape 
being blinded from this, the gate usually causes an after image in the 
same manner so that focusing is still difficult. No brightness above 
the screen level should be visible at the projector position. 

Because of the optical leverage of such a short focal length as re- 
quired for 8-mm use, the misalignment of an objective and finder in a 
camera may easily result in an error of pointing in excess of the usually 
permissible 5 per cent. The widely used negative finder (reverse 
Galilean) allows eye parallax also, not to mention the effects of the 
diffused edge of the mask. Therefore, it is better not to use this 
finder on some cameras. The so-called positive or erecting telescope 
type of finder and the single-lens reflex type such as the Cine-Flex or 
Arriflex offer a solution in principle, but as yet, not at low cost. After 
the turret has been rotated, an error of positioning will also destroy 
the accuracy of a finder, especially if the lens seats are not indexed 
identically for each lens. 

The use of optical rectification to substitute for a methanical inter- 
mittent has many advantages, and a few systems 1 have been devised 
which produce reasonably good pictures. However, to date these 
systems suffer from a poor illumination output because of the re- 
stricted relative aperture and long focal length required of the lenses. 


Perhaps the most advanced and finished work of 8-mm equipment 
may be said to be at present on the mechanical structure of 8-mm 
cameras and projectors; yet some contemporary engineering shows a 
peculiar disregard for the basic requirements; for example, the careful 
finishing of a gear while at the same time misaligning a lens. The 


basic concept behind such construction is no doubt related to the con- 
sumer attitude as conceived by the designing group. 

That the consumer will ultimately choose the equipment giving 
better performance whether it be judged on the basis of cost, quality, 
or convenience cannot be denied without denying the worth of much 
in the way of engineering. On such a basis, it is difficult to see pre- 
cisely how an image either out of focus or improperly pointed can be 
compensated for by an unseen gear built for a hundred years' life. 

The number of lines-per-millimeter resolution required of present- 
day equipment is virtually a definition of the requirements of lens 
mounting as it concerns camera and projector construction. An 
interpretation of the optical requirements of an 8-mm piece of equip- 
ment must be viewed with the thought that it is in reality a projection 
microscope with an intermittent movement working almost as a 
micromanipulator, except at high speed. The projection of each error 
may be said to be roughly doubled, if the same error is made in both 
the camera and projector, and even tripled if a printing is involved. 
The skewing of a lens axis with regard to the correct axis determined 
by the frame center, the mispointing of a lens, or incorrect alignment 
of finder and camera axes may all cause unsatisfactory images whether 
soft or sharp but with cutoff heads of subjects, or both. 

If the camera has one interchangeable lens, or a turret of several, 
the lens-mount problem exists much more than with a single fixed 
lens. It seems like a platitude to say that with interchangeable lenses 
the cameras and lenses must have threads, clearances, and faces to 
match. At present, no standard for 8-mm camera lens mounts exists 
comparable to the C mount for 16-mm cameras. Between the differ- 
ent lens positions of the turret of some makes of cameras one lens will 
focus from 6 to 8 feet to nearly infinity when left at a 12-foot setting. 
This naturally causes disfavor for high relative aperture lenses and 
complicates their purchase. Likewise, in the design of a turret, the 
mechanical conditions may, in another way, limit the performance 
more subtly by restricting the lens designer as to the back clearance 
permissible. If such a problem is posed, the only alternative is the 
use of inverted telephoto types with the added expense of adding and 
subtracting power. Better than a dodge is the use of a turret design 
that allows a maximum of back clearance in depth and diameter. 
The use -of a reflex shutter for single-lens nonparallax focus while 
photographing limits the back clearance, but for a good reason at least. 

A projector faces much the same type of problem. Without a 

446 LEWIS Vol 49, No. 5 

large-diameter lens barrel and adequate back clearance, either wide- 
angle or high-aperture lenses will be restricted if not ruled out. This 
requires a considerable dislocation of mechanism designs. In many 
cases, as the increase of barrel diameter separates the sprockets, and 
likewise causes the location of the pull-down system to be further re- 
moved from the optical axis, there is thrown in more of the effects of 
film shrinkage and the like to add to the problem of unsteadiness. If 
the intermittent is placed on the side away from the lens, the gate will 
best open toward the lens which makes it more difficult to remove and 
replace a field-flattener lens or any other optical system near the film. 

The gate assembly of a projector must also contend with splices, 
some of which will be very bulky and throw the picture out of focus 
visibly when passing. This may be minimized with articulated 
springing so that the splice does not jar the gate open in one frame to 
remain so until the splice passes completely through. An apparently 
classical debate seems to exist in engineering circles concerning the 
use of side tension to position the film in the gate. The head or basic 
mechanism of the projector is then a single unit of design, wherein the 
image requirements set the lens design, which design in turn usually 
determines the illumination-system character. This combination of 
optical systems sets the clearances and diameters to be considered, 
thereby setting the minimum sprocket spacings, how the gate will be 
opened for threading, the location of the intermittent (or the optical 
axis to intermittent separation by frames), and also the shape and 
position limits of the shutter. In short, the head is fitted to the 
optics to obtain the best performance. 

The problem is virtually the same for a camera, with, for example, 
the substitution of a magazine in the place of the condensers. 

In essence, the determination of the resolution and illumination de- 
sired as an end product define conditions of performance which vir- 
tually decide the design conditions of the structure. The desirability 
of a synthesis of design quantities need not be pointed out as opposed 
to the empirical and, therefore, ultimately wasteful system of design- 
ing portions of the head independently to be fitted into a final as- 
sembly. Because of the more widespread knowledge of mechanical 
rather than optical problems, the usual result of the latter method is 
the incorporation of identical and limited lens systems in design after 
design so that, if the mechanical structure were improved, it would 
not show very well. 

The point of translation between optical requirements and the 

Nov. 1947 


performance of the pull-down system is the definition of dynamic reso- 
lution; namely, the exact nature of the equation connecting the reso- 
lution of a single still image and the projection of a succession as in a 
motion picture as seen by the eye. When this relationship is clearly 
defined, the tolerances of the entire system may be defined and toler- 
ances balanced for manufacturing use. The factor between still and 
dynamic quantities is also determined by the film-perforation accuracy 
as well as persistence of vision. 

With the film held to a tolerance of 0.0005 inch (0.013 mm) 
between successive perforations, the low limit of resolution between 
successive frames, superimposed by persistent retinal images is 
roughly the reciprocal of the tolerance (Vo.ow = 77), or 77 lines per 
millimeter. If the total of the tolerance is considered, however, the 
low limit sinks to 38 lines per millimeter. If the sidesway is controlled 
by the perforation, 100 to 50 lines per millimeter may be expected as 
a low limit. The width of the film, if allowed to control sidesway, 
will yield but 12 lines per millimeter, as opposed to guiding by the 
perforation side only, which gives about at the least 10 lines per milli- 
meter, corresponding to the same tolerances for the aperture center line. 

Considering the tolerances assigned to 8-mm film, the stacking of 
components enables the following estimation of the errors between 
successive frames as shown in Table 2. 


Lowest One-Sided 

Permissible Error Only 

Horizontal line resolution 21 43 

Vertical line resolution 10 20 

That such low limits are rarely reached is a matter of experience and 
empirical trial. It seems highly improbable that the vertical lines 
will be so poorly resolved, as the gate is normally several frames long 
such that the worst error drops easily to one fourth. 

If several teeth are brought to bear on the film, the errors begin to 
reduce within the limits permitted of shrinkages such that between 
successive frames the stacked tolerance may be reduced considerably, 
depending somewhat on the pull-down teeth shapes and spaces. If 
the film shrinkage is within the usual limits and uniformly the same in 
amount throughout the entire length of the film, no effect should be 
readily discernible ; if, however, a 2 per cent differential exists between 

448 LEWIS Vol 49, No. 5 

frames, a low of 13-line-per-millimeter resolution may occur. Such 
a shrinkage seems very improbable. 

If the conditions are such that film perforations permit resolution 
equivalent to the optical system, the remaining link is the film- 
handling system. If the film-advance system is of a single-cycle type, 
has no nonrepetitive errors between frames, and if the film is not 
moved during projection, the picture should be asbolutely steady. 
However, inasmuch as many intermittents are mechanically magnify- 
ing the cam as translated into film motion, a variation between frames 
or cycles of lubricant film thickness of one half a thousandth of an 
inch will most likely result in a resolution drop to about 20 lines per 
millimeter. Under such conditions a reduction of leverage would 
appear advisable. At such tolerances, the use of a sprocket inter- 
mittent of six or eight teeth appears impracticable because of the 
probable six- to eightfold magnification of manufacturing errors in 
addition to those just cited. The optical intermittent of the com- 
pensating and continuous moving-film type is unlikely to have a very 
good opportunity to compete in so far as resolution is concerned, as 
it must be driven by a sprocket or similar construction. 

As the resolution of an 8-mm piece of equipment is definitely a 
quantity related to its design, the designation of the means of measure 
of such is of value. The resolution of the objective may well be 
handled by tests similar to the standard 16-mm projection-lens test. 
The film constitutes no problem. The dynamic resolution, on the 
other hand, constitutes a problem depending upon the conception of 
persistence of vision. If the eye were a device opening and closing 
its sensitivity rapidly, the problem would be simple, but the smooth 
and individually variable slopes of the curve introduce enough vari- 
ables to confuse the issue. It is suggested that the dynamic resolution 
be tested by normal motion picture projection through a shutter for 
an exposure of between 1 /i to 1 /ie of a second, using a film as screen 
to integrate the image during exposure. A test of this type integrates 
the total performance of the system from subject through camera and 
projector but not including the screen. Utilizing this approach, 
Table 3 was obtained from equipment chosen at random. 

Photography of the screen image completes the resolution chain 
but adds an unnecessary factor. 

The camera film-handling system must be driven at such a speed 
that the exposure does not change visibly during a scene. Also, a 
camera must start and stop within a frame, always dark in the dead 


position. The mechanism other than the governor and intermittent 
may therefore be crude but not so crude that it is noisy and causes 
distraction of the subjects. The speed regulation may be due to a 
governor exerting a drag. Escapement regulation similar to a watch 
has not as yet appeared in cameras, though various electric controls 
are being used for cameras and projectors. 


Vio second 80 (a few lines of 112 L/mm show occasionally, Fig. 1) 

V* second 40 (grain pattern gone entirely, Fig. 2) 

Visual 80 

Among various factors causing image deterioration, the rigidity of 
support of the camera itself is a factor in so far as the consumer is con- 
cerned. A loose tripod thread and a small base on the camera are not 
good aids to secure mounting. The hand-held use of pocket-type 
cameras is also a very important item in this regard. An angular 
movement of Yioo radian may cause a degeneration of resolution down 
to 8 lines per millimeter, if it occurs in a period equal to one frame. 

Numerous aspects enter into such a picture, including the type of 
sight, the trigger-release pressure, and its position, not to mention the 
possible friction of the trigger end. The type of case construction 
enters into the picture from the same tactile sense. A further -factor, 
usually ignored, is the moment of inertia of the moving parts. The 
jerk on starting or stopping can well cause such a loss in sharpness. 


That sound with 8-mm film is desirable cannot be denied; the prob- 
lem is more likely, what is acceptable ? The use of disks or magnetic 
records synchronized with the films is capable of very fine performance 
except for the already well-known troubles of film repair and syn- 
chronization. Sound on film for 8-mm use at present is still a matter 
for laboratory discussion. 

Optical tracks may on the basis of current practice reach as high 
as 3500 cycles if 7000 is accepted as the top for 16 mm. With high- 
resolution stock (microfilm) and precisely controlled conditions, fre- 
quencies as high as 5000 cycles can be prophesied so far as the labo- 
ratory is concerned. Yet, the problems of 8-mm sound on film hinge 
on other factors, yielding to indirect attack only. The noise level in 
8-mm optical sound is so high that, if good frequency response is 



Vol 49, No. 5 

attained, it must again be thrown away by the use of cutoff filters. 
Volume expansion by means of a separate control track has been sug- 
gested. The amount of space available for the sound track or tracks 
is not very adequate, as the picture area must be robbed or the track 



FIG. 1. National Bureau of Standards charts were photographed by an 
8-mm camera, and the film processed and projected upon photosensitive ma- 
terial for a screen. Exposure is made by a shutter in front of the projector. 
These charts were at twice the usual distance such that the line-per-millimeter 
resolution must be doubled. (Top line, 112 lines per millimeter.) Extreme 
care is required in taking such pictures to attain the maximum resolution as il 
fluctuates rapidly. 

Nov. 1947 



placed outside the sprocket holes. This places the sound track in a 
location where developer activity is very likely to produce flutter, 
unless a viscous developer is applied by rollers or the like to avoid the 
effects of the adjacent sprocket hole. The very slow speed of 8-mm 
film even at 24 or 30 frames is such that the construction of film- 
transport systems for sound reproduction becomes very difficult. 


FIG. 2. A Va-second exposure under the conditions of Fig. 1 produces a 
pattern wherein grain is no longer of consequence. If the motion picture 
equipment were perfect and the film perfectly perforated this exposure would 
probably exhibit higher resolution than Fig. 1. 

452 LEWIS 

The 8-mm sound-on-film scanning systems will require better optics 
than the 50-line-per-millimeter slits at present used for 16-mm sound 
on film and also a spacing of more frames between picture and sound. 
(Fifty-two frames would allow direct conversion of much 16-mm 
sound-on-film equipment.) 

At first glance one would suspect that the frequency response pre- 
dicted for 8-mm magnetic sound on film 2 was a very poor condition. 
However, the lack of difficulty with dynamic range and sprocket-hole 
modulation caused by developer products and a very favorable noise 
ratio show very great promise in addition to the advantage of being 
able to record at home without additional processing. The variation 
in flexing of the film at the sprocket holes is no more likely to be a 
problem for magnetic than for optical sound. 


With the advent of sound for 8 mm, and the construction of the pro- 
jectors as a complete unit assembly including the screen, the installa- 
tion of motion picture projectors, along with television sets, is likely 
to become a parallel in sales to the installation of phonographs with 

A complete study of the performance of 8-mm material in terms of 
resolution has not been published and is definitely needed to enable 
adjustment of design to avoid unnecessary losses. Among the vari- 
ous items needed are data on printing resolution, screen resolution, 
maintenance of the focal plane in operation, and more thorough data 
on dynamic resolution and persistence of vision effects. 

There will probably arise a desire for the study of the numerical 
aperture of optical printers, the effects of diffraction in printing, and 
the possibility of higher-resolution color film. 

In the process of realizing the above, it seems safe to prophesy that 
the future will see the increased use of basic-design quantities derived 
from laboratory research rather than the traditional and strictly draw- 
ing-board approach. 


The writer wishes to thank Eugene L. Perrine for his help in making 
the tests upon which this paper is based. 


1 BACK, F. G., AND EHRENHAFT, F.: "A Non-Intermittent Motion Picture 
Projector", /. Soc. Mot. Pict. Eng., 34, 2, (Feb. 1940), p. 212. 

2 CAMRAS, MARVIN: "Magnetic Sound for Motion Pictures", J. Soc. Mot. 
Pict. Eng., 48, 1 (Jan. 1947), p. 14. 


Summary. This paper outlines the design objectives involved in the production 
of a precision-made 8-mm projector, consideration being given to sales attractiveness 
in styling, distinctiveness, and simplicity of appearance; problems in film handling, 
simplification of functions, minimizing operational controls, illumination efficiency 
as correlated with film movement and optics; temperature control, and production cost 
economy in parts and assemblies. 

Exhibiting equipment for 8-mm moving picture film became pre- 
dominantly domestic because of its use being confined primarily to 
the individual or the amateur enthusiast. 

Under these circumstances, it seemed imperative that a distinct 
departure be made from the orthodox style of projector design which 
is prevalent in 16-mm equipment. This probably was the first in- 
stance of a machine of this type being accepted as a definite part of 
the home living room. 

It seemed evident therefore that the design of an 8-mm projector, 
in order to be acceptable in these surroundings, should embrace sim- 
plicity of design, attractive proportions, and distinctiveness of style; 
also it should be constructed so as to conceal moving mechanism as 
much as practicable and to minimize operational controls. 

These were the design objectives outlined for the 8-mm projector, 
whose components and functions are to be described. 


A universal motor was selected for this projector to provide either 
alternating- or direct-current operation, because of the economy in 
omission of the auxiliary equipment for operation and the preference 
for its high starting torque qualities. The projector load seemed suf- 
ficiently constant to overlook resultant speed changes and a rheostat 
control as provided, satisfactorily compensated for line-voltage and 
motor variations. Inherent noises in this type of motor are minimized 
by the control of the quality of the brushes and the commutator sur- 
face. The motor brushes are readily accessible for replacement. 

The motor was so located as to provide adequate cooling by its 

* Presented Apr. 25, 1947, at the SMPE Convention in Chicago. 
** Ampro Corporation, Chicago, 111. 


454 , MORGAN Voi 49, No. 5 

shell exposure to open air and induced air currents through it leading 
directly toward the air intake of a housed fan mounted upon an ex- 
tension of the motor shaft. The motor mounting is of the flange type 
and the motor is readily interchangeable in servicing without dis- 
turbing any of the mechanism. 

The projector switch is the main control and the projection-lamp 
switch is fed through it, thus assuring lamp ventilation at all times 
except for possible motor failure or stalled mechanism. A motor-re- 
versing switch is provided for either . rewinding or reverse projec- 
tion, and the plane of action of its handle is purposely positioned at 
right angle's to the handle of the main switch for touch distinction in 
the dark. Reverse projection is obtained when the projector is run- 
ning forward by merely throwing the motor switch to "reverse". 
These switches are closely and conveniently grouped in the base near 
the rheostat and the line receptacle for ease of operation, short leads, 
and connections. 

An uncommon feature is the miniature threading lamp. It is pro- 
vided without a control switch and so wired as to be "on" whenever 
the projection lamp is "off". A threading lamp effectively located, 
but also necessitating a series resistor, seemed preferable to an ob- 
viously cheaper installation of a larger 110-volt-type lamp ineffec- 
tively located elsewhere. 

A simple economical and effective drive is provided from the motor 
to the main drive shaft by the use of a molded V belt and pulleys. 
This belt is also readily accessible for replacement if necessary. 

The main drive shaft is provided with a simple friction disk clutch, 
both driving and driven members being zinc die castings. The driv- 
ing member is designed to serve as a clutch member, a pulley, and a 
flywheel. Disengagement of this clutch stops all film transport mecha- 
nism for the express purpose of exhibiting still pictures. 

This main shaft serves to drive two individual gear trains; one, the 
safety shutter and intermittent mechanism, and the other, the two 
12-tooth film sprockets. Both upper and lower sprocket shafts in 
turn serve to drive both upper and lower reel spindles, respectively, 
by means of coiled spring wire belts. 

The gearing is so arranged as to permit a higher velocity in rewind- 
ing, which is simply accomplished by threading the film directly 
from the take-up to the feed reel, by throwing the motor switch to 
"reverse" and the main switch to "on". The rewinding of 400 feet of 
film requires about ! 1 / 2 minutes. 



The shuttle is of the simple stamped lever type having its pivoted 
end so mounted upon a flat bronze spring as to permit of free vertical 
and horizontal oscillations alternately. This spring is anchored upon 
a bar having liberal vertical and minute lateral adjustments. On 
the opposite end of the shuttle is formed a single claw for film engage- 

The vertical adjustment provides film picture framing with a fixed 
aperture while the lateral adjustment permits claw-stroke regulation 
when the shuttle is maintained in working contact with its actuating 



FIG. 1. Cam-unit assembly, exploded. 

cam. This contact is made through means of a hardened and ground- 
steel follower button fastened to the shuttle. A similar button is 
used as the contact means between the shuttle and the vertical cam- 
follower spring. A Bakelite button on the shuttle is used as the lat- 
eral cam follower. 

The cam-unit assembly (Fig. 1), mounted upon a stud, consists of 
a Bakelite gear (1) ; staked to an oilite bushed steel hub (2) ; mount- 
ing the vertical cam (3) ; a spacer washer (4) ; a three-bladed shutter 
(5); from which is extruded the lateral cam surface (6), which is cir- 
cular in form and operates axially. This cam surface is oriented in 
timed relationship with the shutter travel blade. These three mem- 
bers are in turn keyed with a pin (7) in a predetermined time rela- 
tionship and locked together with a special nut (8). 

456 MORGAN Vol 49, No. 5 

On the vertical cam the film-transport sector is represented by a 
curve subtending an arc of 24 degrees or the equivalent of a 14-to-l 
movement. This enables one to use a shutter having three blades of 
only 35 degrees each and results in an extended light-exposure period 
amounting to 70.8 per cent of the picture cycle. This shutter is made 
of exceptionally heavy gauge steel in order to effect a flywheel 

Adequate lubrication facilities, clearly indicated, are provided for 
all moving parts with particular consideration being given the inter- 
mittent unit. 


Both feed and take-up spindles are driven by means of coiled spring 
wire belts and pulleys concealed from view within gracefully shaped 
die-cast arms designed as integral portions of the mechanism head and 
having removable covers at their rear thus allowing full access for 
belt replacement. The projector-carrying handle was also cast as a 
part of the upper arm for cost economy. 

These arms were so designed and the spindles therein so spaced as 
to accommodate film reels of 400 feet capacity and thus provide a 
maximum continuous exhibition of 33 minutes at normal projection. 

The take-up reel during forward projection having a gradually 
diminishing angular velocity clockwise, necessitates a device which 
will compensate for this change while driving the spindle. Except for 
the slight drag necessary to proper film-rewinding tension, this spindle 
must also be free to rotate in a counterclockwise direction during the 
rewinding operation. 

Such a device was developed to perform these functions auto- 
matically and thus eliminate the necessity for any additional manual 
controls. Briefly, this device operates in reverse to the principle in- 
volved in a capstan where the amount of pull by the drum upon the 
fixed end of the medium is dependent upon the number of convolu- 
tions thereon and the degree of tension exerted at the free end of the 
medium. The reverse condition in this device is that neither the 
drum or spindle becomes the driven member through means of the 
medium, in this case a coiled spring attached at one end to a driven 
pulley and definite frictional drag applied at its free end. Under 
static conditions rotational impetus directly applied to the reel 
spindle in either direction has no influence whatever upon the winding 
or unwinding of the coiled spring which has relaxed normally and thus 
freed itself from the spindle. 


All slippage accompanying velocity differentiation is absorbed 
within this device and therefore none is demanded of the driving 
belt or pulleys. 


The aperture plate is not provided with the customary film-guiding 
channel but has only a slightly depressed track for picture-protecting 
clearance. A plate of this type has less affinity for the accumulation 
of dust particles. 

Three fixed film-edge guides on one side of the aperture are opposed 
by two film-edge-pressure springs on the other. Because of the edge 
curvature manifest in 8-mm film, the edge guides above and below the 
aperture are positioned at a greater distance laterally from the optical 
axis than the central guide directly opposite the aperture which ac- 
curately aligns the film with this axis. The edge-pressure springs are 
located symmetrically above and below the horizontal plane through 
the optical axis. The action of opening the film gate automatically 
retracts the edge-pressure springs for freedom in threading film be- 
tween them and the edge guides. 

The film gate is opened by means of a cam lever conveniently lo- 
cated upon a sliding lens holder in which is retained the film-pressure 
shoe. This shoe, although spring-floated and accurately retained in 
operating position, is free to be withdrawn easily for cleaning or re- 
placement only when the gate is open. The stamped H spring urging 
this shoe is also readily removed for replacement without tools. 

Shoe pressure on the film is controlled by an adjustable stop 
mounted upon the lens holder. Free access for properly cleaning the 
picture aperture is possible when the objective lens and the pressure 
shoe are removed. 


The original design of this projector comprised a system having as 
its light source a 500-watt, T-10, 115-volt medium prefocused base 
lamp. Subsequently, however, the adaptation of a 750-watt, T-12 
lamp and modified optics for same within the limits of the original 
lamphouse, included a 1-inch E.F., coated, //1. 6 objective lens, a 
single element, aspheric-convex condenser 22-mm outside diameter 
and a rhodium-surfaced reflector of 1 inch radius of curvature. 

Illumination uniformity and output test results for this system us- 
ing 750-watt lamps and a 16-mm standard test chart on a 40-inch 
screen were as follows: 



Vol 49, No. 5 

The linear aperture magnification in the above tests for an 8-mm 
aperture of 0.172 X 0.129 inch is 232 to 1, while the linear aperture 
magnification in similar tests for a 16-mm aperture of 0.380 X 0.284 
inch is only 105 to 1. 

Objective lens focusing is accomplished in an orthodox manner of 
spring-ball engagement with the helical scoring provided on the lens 

The projection-lamp socket is provided with a mounting unit hav- 
ing a simple adjustment screw for centering the lamp filament later- 
ally. This screw may be adjusted externally of the lamphouse with a 







7.0 7.2 










6.9 9-9 



FIG. 2. 


Averages of 6 lamps in foot- 
candles. Average drop-off at 
corners = 17 per cent. 

Averages of 6 lamps in foot- 
candles. Total lumens out- 
put = 62.75. 


coin or similar instrument. The condenser unit is accessible when the 
lamphouse cover has been removed and is easily withdrawn for clean- 
ing or replacement of the element without tools. The reflector is ad- 
justable along the optical axis only and may be removed easily for 

Having the ventilating fan directly mounted on the motor shaft 
presented a very interesting problem in the design of the fan-housing 
scroll where the necessity for equal air delivery in either forward or 
reverse fan rotation was apparent. A unique solution of this problem 
thus resulted. 

The usual procedure in designing the spiral-scroll curvature for a 
single-direction fan housing is a common problem. Our problem ne- 
cessitated the development of both a right- and left-hand spiral scroll 
from a common cutoff. 



These spiral curves (1), Figs. 3A and 3B, obviously intersect at 
a point (2), upon an extended line (3), drawn between the center (4) 
of the fan, and the common cutoff (5) . This fact, therefore, definitely 
set the angular limit to which we could develop fixed-scroll curves 
moving inwardly or toward the center of development. It was ob- 
vious that the sectors necessary to the completion of each scroll be- 
tween this point of intersection and the cutoff were similar. It eventu- 
ally developed that one such sector alone would serve to complete 

FIG. 3. 

Vane position with fan rotating in 
forward position. 

Vane position with fan rotating in 
reverse direction. 

either scroll when constructed as a vane (6), whose contour approxi- 
mated that of the normal sector and which was free to travel between 
the limits of these two positions. Either of these positions would be 
automatically determined by and dependent upon the directional im- 
petus imparted to the vane by the air discharge from the fan (7) itself. 

The limited angular spiral development under the above circum- 
stances is recognized. 

Anemometer tests for air-discharge volume and velocities, through 
a special stack, taken 12 inches above the top of the lamp for both for- 
ward and reverse fan directions and with the projector operating at 
16 frames per second are shown in Table 1. 





Table 1 

Air Discharge 
Feet per Minute 


Vol 49, No. 5 

Cubic Feet 



Highly effective aperture cooling was accomplished by directing 
induced cool outside air currents across the back of the aperture 

FIG. 4. Air intake ports for aperture cooling. 

through ports indicated in Fig. 4, leading directly into a narrow heat- 
insulating chamber between the aperture and the lamp compartment 
and connected with the fan intake. 

The lamphouse-cover unit includes lamp shielding providing 
double air-space insulation protection between lamp and external sur- 

A grille as part of the lamphouse cover proper was designed to re- 
duce ceiling illumination with a minimum impediment to discharging 


air currents. Portions of the grille-design pattern projected onto the 
body of the lamphouse dually served in heat dissipation and in styl- 
ing treatment. 

Temperature tests for cooling efficiency at critical locations are 
shown in Table 2. 


Forward Reverse 

Location of Thermocouple Degrees Degrees 

Aperture, 7 /i6-inch radius from its center ... 53 . 5 54 

Lamphouse, left side* 83 . 84. 5 

Lamphouse, right side* 59 . 61.0 

Lamphouse, rear*. 61.0 64.0 

* Thermocouples located external of lamphouse, in optical plane and directly 
opposite center of light source. 

Temperature readings are centigrade scale. 

Duration of "forward" tests were 45 minutes while "reverse" tests, which fol- 
lowed, were 15 minutes. 


The tilt mechanism on this projector provides 15-degree upward 
and 5-degree downward adjustment of its optical axis. This is ac- 
complished by the provision of substantial support in the main base of 
the projector for a sturdy pivot upon which the main projector head 
may rotate. A leg extension, forming a rigid part of the head, and 
channeled to conceal wiring passing around this pivot, projects down 
into the base chamber. The leg's lower extremity is provided with a 
slot designed to accept and fit a stud forming a rigid part of a special 
elongated die-cast nut. Because of its intimate contact with this leg, 
the nut is restrained from rotation. A shaft substantially journaled 
in the projection base and restrained from axial travel has a threaded 
portion projector within the base, designed to fit and support the nut. 

Manual rotation of this shaft thus indirectly imparts a rotating or 
tilting movement to the projector head by the resultant axial move- 
ment of this special nut-stud component. 

Obviously this device is self-locking and permits of accurate and 
effective adjustment. Care was exercised to locate the main pivot 
directly below the mid-point between the two extreme positions of the 
center of gravity of the projector mechanism. The net result of this 
effort was uniform ease of adjustment. 


Manufacturing economy was realized in resorting to the use of die 
castings. Outside of a few critical parts, these castings required little 
more than simple drilling and tapping operations. An obvious ad- 
vantage was the reproduction of neat, clean-cut, graceful shapes 
and forms, some much too intricate to contemplate otherwise, 
especially in view of the necessity for mass production. 

Two typical examples of machining and die cost economies ef- 
fected are cited here. First, rectangular openings in both the top and 
bottom of the mechanism head were required for the clearance of 
spring belts. These openings were inaccessible for direct machining. 
A slight change was made in the shape of a moving die core so the 
plane of its top surface would be coincident with the wall of the die 
cavity. This resulted in the simple coring of the desired opening 
which only necessitates removal of a slight flash in the raw casting. 
Second, very narrow gib slots in one casting were necessary for the 
guidance and retention of the condenser-lens-holder casting. This 
was quite impracticable to produce in the die with a sliding core be- 
cause of its taper, depth, and narrowness. Straight coring from one 
side of this casting at the upper and lower extremities and similar 
straight coring from the opposite side of the casting in the center por- 
tion produced the desired effect of retaining gibs on both sides. 


The general over-all dimensions of the projector are 9V4 inches long 
X 6 13 /ie inches wide X 13 7 /s inches high. The outside dimensions 
of the carrying case, exclusive of hardware projections, are 11 inches 
long X 8 3 /s inches wide X 14 7 /s inches high. 


The net weight of the projector complete is 13.9 pounds. The 
gross weight of projector and carrying case including standard acces- 
sories is 22.2 pounds. 


Grateful appreciation is hereby acknowledged for the helpful dis- 
cussions with A. S. Dearborn and T. R. Neesley of this company 
upon the subject of this paper. 


Summary. The first commercial 8-mm sound projector has been introduced 
with the sound on a disk running at 33 l /z revolutions per minute. An automatic- 
synchronization method is used, and the turntable and the projector are not mechani- 
cally connected. Eight-millimeter sound films for use with the machine are available . 

The first 8-mm commercial sound projector has been introduced to 
the market under the trade name of Movie-Sound 8. Sound on disk is 
the method used in obtaining the sound, and while sound on disk is 
certainly not new, there are two features used with the unit which we 
believe are new and which makes sound for 8 mm practicable. 

The two things are, first, an automatic method of starting the film 
and disk in synchronism is used. This is done by recording on the 
disk, a one-thousand-cycle tone in synchronism with the synchroniza- 
tion mark on the picture. The projector is threaded with the syn- 
chronization mark in the gate. The pickup needle is set down on the 
revolving record. The one- thousand-cycle tone is picked up, and this 
operates a selective relay which starts the projector motor, and starts 
the two separate units in synchronism. It is true that a motor cannot 
start instantaneously but on a lightweight projector such as an 8- or 
16-mm projector it will start ?lmost instantaneously, but more im- 
portant, the starting characteristic is quite uniform. Therefore it is 
possible to place the synchronization mark on the film so as to com- 
pensate for the loss in starting time. A special standardized leader 
has been made up and this leader is spliced on the beginning of any 
film which is to be printed and re-recorded for use on the Movie- 
Sound 8. This leader contains the 1000-cycle starting tone and the 
synchronization mark in their correct positions to make the sound and 
picture start in synchronism. The turntable and the projector motor 
are both synchronous motors, and once they are started in synchro- 
nism, they will run indefinitely in perfect step. 

Second, the 8-mm picture prints which are used with this projec- 
tor, when printed from a standard 24-frame-per-second sound film, are 
skip-frame-printed to project at 16 frames per second. This means, we 

* Presented Apr. 25, 1947, at the SMPE Convention in Chicago. 
** Continental Products Corporation, 1103 E. 15 St., Kansas City, Mo. 


464 THOMPSON Vol 49, No. 5 

have a picture running at one speed, 16 frames per second, and the 
sound track on the disk at 24 frames per second, running in synchro- 
nism. This 16-frame-per-second speed was chosen for a number of 

It allowed the use of a standard projector speed at 16 frames per 
second which is the standard speed for taking and showing of 8-mm 
films. Since many 8-mm cameras will not run at any other speed, it 
was felt desirable to make the unit project at the standard speed. 

It will save film for the amateur. Some amateurs will want to 
make records to use with their own films, and a recording unit will be 

FIG. 1. Movie-Sound-8 unit set up for use. 

available soon which will make records with the one-thousand-cycle 
starting tone. In that case it is desirable that 16 frames per second 
be used. Experience has shown that amateurs object to photograph- 
ing 24 frames per second, as they consider that it increases their film 
cost by 50 per cent. 

A saving in film cost can be made in printing library films to be 
used with the projector. 

A one-speed synchronous projector is less expensive to build than a 
two-speed synchronous projector. 

Amateurs who want to add a synchronous motor to their present 
projectors can do so much more easily if they do not have to think 
about making it run at two speeds. Such a converted projector will 
work with the Movie-Sound-8 system. 

It is believed that there are a number of advantages to sound on 

Nov. 1947 



disk to be used for amateur home use only, and at present, at least, 8- 
mm is the amateur movie film. Some of these advantages are de- 
scribed in the following paragraphs. 

Almost any quality of sound 
desired by the amateur can be 
obtained with disks. The art of 
making electrical transcriptions 
has been fairly well standardized, 
and today this method of mak- 
ing delayed broadcasts is used 
by every radio station in the 

Excellent quality can be put 
on the disk after which it can be 
played on any type of playback 
with various degrees of quality. 
If the amateur wants the best 
quality of sound with his pictures 
there is no reason why large 
speakers, powerful amplifiers, 
and high-quality pickups can- 
not be used. On the other 
hand, inexpensive units giving 
quality equal to the average home 
radio can be used. 

The Movie-Sound-8 amplifier has a frequency response from 100 to 
6500 cycles essentially flat. The crystal pickup and the medium-size 
speaker necessary for a one-case job naturally limit the quality some- 
what. The quality of the SSYs-revolution-per-minute transcriptions 
used for the sound track is naturally governed by the quality of the 
original film recording. 

If the amateur obtains his sound from a disk, a turntable is part of 
the machine and this turntable can run at two speeds. This allows 
him to play regular 78-revolution-per-minute records, and the unit 
can be used as a portable phonograph or play records as background 
music for his own pictures. Since phonograph records are the only 
common source of recorded music available to the amateur, this is an 

If the amateur desires to add narration to his own pictures, a record 
can be made with the automatic start tone and the only material cost 

FIG. 2. 

Complete unit packs into one 

466 THOMPSON Vol 49, No. 5 

will be that of the record. There is no sound film to buy, no process- 
ing, no prints. His original film will be the only print he needs. If 
he desires to change some of his scenes after he has made the sound, 
he has only to take out the scenes he does not want and add an equal 
footage of a better scene. Since amateurs do not always secure the 
scene they desire the first time, this should be an advantage. 

Amateurs are used to handling records. Anyone who can thread a 
projector and operate a phonograph can operate the Movie-Sound 8. 

There is no reason why the sound unit cannot be built into a radio- 
phonograph combination which will work in conjunction with the 8- 
mm projector, if the Movie-Sound-8 system is used. 

The radio-phonograph manufacturer will have to make two 
changes. A two-speed synchronous turntable is necessary, and the 
automatic-relay device will have to be incorporated into the amplifier 
circuit. With such an arrangement a synchronous projector can be 
plugged into an ordinary radio set, and the amateur can have sound 
movies at home with no additional expense except that of a synchro- 
nous projector. 

The Movie-Sound-8 projector is a special unit* driven by means of a 
chain drive from a 1 / 40 -horsepower capacitor start-and-run 3600- 
revolution-per-minute synchronous motor. The projector runs at 
exactly 16 frames per second when operated on 110-volt, 60-cycle, 
alternating current. Any 8-mm projector driven in a similar manner 
could be used with equal success. An ordinary projector driven by 
a variable-speed motor will not, of course, work. 

The amplifier** is a straightforward alternating-current amplifier, 
such as is used in a good-quality record player with the selective-fre- 
quency relay built into it. The speaker is muted, and is turned on at 
the same time the projector is started. There is a pilot lamp on the 
turntable which automatically goes out when the projector starts. A 
two-speed synchronous turntable is used in the machine. A speed of 
33 Vs is used for motion picture projection, and a speed of 78 revolu- 
tions per minute for phonograph records. An ordinary silent slide 
projector can also be used with the unit for showing sound slide films. 
The amplifier and the turntable motor are both readily removed in 
case service is needed, and any qualified radio-repair shop can easily 
make any needed repairs. A selected number of 8-mm sound library 

* Built by Eastman Kodak Company especially for the Continental Products 
Corporation, known as the Kodascope- Eight CPC. 
** Built by the Wilcox Electric Company. 


films have been released for use with the machine by Castle Films, a 
Division of United World Films. Any 35-mm or 16-mm sound film can 
be printed and re-recorded for use with the Movie-Sound-8 unit. In 
making the negative from which the picture prints are to be made, the 
skip-frame method is used so that all prints resulting from such a 
negative will run at 16 frames per second. The sound is recorded on 
a disk, and pressings of these records are made in the usual manner. 


MR. A. SHAPIRO : Does the record rotate at 33 1 / 3 or 78 when used with 8 mm? 

MR. LLOYD THOMPSON: 33 Va revolutions per minute. 

MR. SMITH : How many feet can you get on one record? 

MR. THOMPSON: One ordinary reel by using a 12-inch disk. If you want to 
use a 16-inch disk it is possible to get two reels on it, which would be equal to 
2000 feet of 35-mm sound film. 

MR. REED: How many frames do you lose in getting started? In other 
words, how many frames does it require to start the projector before lip synchroni- 

MR. THOMPSON : I cannot tell you how many frames are actually lost in getting 
the projector started because I have never counted them that way. However the 
starting characteristic is quite uniform. 

MR. JORGENSON : You mentioned that amateurs could cut out portions that were 
unsatisfactory and splice in better scenes. How do you manage to synchronize 
words or actions in that case? 

MR. THOMPSON: We are not talking about a picture which an amateur might 
make and record. If you cut out one sequence and put in another you would 
have to put in the same number of frames that you took out. 

MR. JORGENSON: How about the motion of the lips? 

MR. THOMPSON : We are still talking about a picture which an amateur might 
make and record. It is not intended to be used for lip synchronization because it 
would have to be faked. I do not know of any practical way for amateurs to do 
lip synchronization with this or any other system. Lip synchronization on li- 
brary films will of course play in synchronism on the Movie-Sound 8 if they were 
made in synchronization in the original production. 



Summary. This paper describes a new combination sunshade and filter holder 
which is designed for use on almost any 16- or 8-mm camera. The device is so de- 
signed that it not only acts as an adequate light- and sunshade for the camera lens, but 
by employing a series of slides it will accept gelatin or cemented filters, diffusion disks, 
gauzes, and pola-screens in various sizes. 

Recently the introduction of a new color-film emulsion by one of 
the largest film manufacturers in this country again brought to the 
attention of the 16- and the 8-mm camera users the fact that their 
cameras were not adequately provided with a means whereby a vari- 
ety of filters could be used. This new color film, in many cases, re- 
quired the use of one or two filters for correct color balance. Imme- 
diately the problem arose of how these filters could be supplied to 
the amateur and semiprofessional cinematographer in sizes which 
could be used on most all of the 16- and 8-mm cameras. The desire 
to use these special filters in the above-mentioned case, however, was 
not the only time that this problem has presented itself. Up to the 
present time, it has been almost impossible for the cinematographer 
who normally uses 16- and 8-mm cameras to equip himself with any 
professional type of sunshade and filter holder which would permit 
him to use standard filters, diffusion disks, pola-screens, and gauzes. 
This was especially true in connection with filters, because of the fact 
that they are normally supplied in a variety of standard sizes. 

The combination sunshade and filter holder is simply and sturdily 
constructed and is similar in design to the well-known and standard 
"matte boxes" which have been used on professional 35-mm cameras 
for a number of years (see Fig. 1) . By employing a series of removable 
slides the device will accept any filter, diffusion disk, pola-screen, or 
gauze in the following sizes: 3-inch square, 2 1 /2-mch round, 2-inch 
square, and the Kodak Series VI filters. In addition to the numerous 
filter sizes mentioned above, the device will also accept any gelatin 
filter, such as Wratten filters, by means of a specially constructed 
slide. This slide will hold up to four of these gelatin filters which may 
be cut easily and placed in the slide. The slide is then inserted into a 

* Presented Apr. 25, 1947, at the SMPE Convention in Chicago. 
** Bardwell and McAlister, Inc., Hollywood, Calif. 




special slot provided which positions it directly in front of the camera 
lens (see Fig. 2). 

This combination sunshade and filter holder is a universal device 
and may be used on almost any 16- or 8-mm camera. It is not nec- 
essary to drill holes or alter the camera in any way, and the device is 
so constructed that it is adjustable in all planes and may be correctly 
centered in front of the camera lens. This is accomplished by provid- 
ing a camera base which will fit on any amateur or professional tripod. 

FIG. 1. 

The camera is mounted on this base which has a dovetail machined in 
the casting on the forward edge. The rod assembly which holds the 
sunshade and filter holder is mounted on this dovetail and can be 
moved to the right or left as far as desired. All other adjustment 
movements are accomplished by moving the device up and down on 
the slide casting and forward and backward on the slide rods. There 
are a few cameras, however, which are constructed with the view- 
finder positioned very close to the photographing lens. This, of course, 
was done purposely in order to eliminate as much parallax between 
the photographing lens and the view finder as possible. In these few 
cases, however, it is only necessary to attach to the sunshade and filter 



holder an auxiliary view finder because of the fact that when the 
sunshade and filter holder are used it obstructs the camera view finder. 
The device is provided with a boss on the left side upon which an aux- 
iliary view finder may be attached in order to eliminate this difficulty. 
In such cases an auxiliary view finder can be supplied and is adjustable 
for parallax and also is provided with a series of mattes to match the 
field of view of the normally used lenses. 






(E) 3" SQ. GAUZES 


(B) ANY 1-5/8" DIA. FILTER 





FIG. 2. 

(A) 2-1/2" RD. FILTERS 



Because of the fact that wide-angle lenses are becoming more and 
more popular for use on 16-mm cameras, the device was designed with 
an angle of acceptance wide enough to permit its use with a 15-mm 
lens on a 16-mm camera. The various combinations of filters, gauzes, 
and diffusion disks which may be used in the device at one time are 
numerous. It also permits the use of the standard 3-inch square and 
2-inch square graduate filters which up until this time have been ex- 
tremely difficult to use in conjunction with 16- and 8-mm cameras. 
Besides having the ability to accept a great number of filters of various 
sizes, the device is finished with a black flock material on the inside of 
the shade which acts as an excellent light deflector. The device may 
be used when the camera is hand-held, or in conjunction with a tripod. 
It is recommended, however, that a tripod be used wherever possible. 



Summary. The operation of a spring-driven motion picture camera is seriously 
affected if a mechanical tachometer is used to measure the speed. Equipment has 
been developed and made to give a direct reading of camera speed utilizing a light beam 
as the connecting link. 

The measurement of the picture frequency of cinematographic ap- 
paratus is one of the more important aspects of its manufacture. If 
the apparatus is powered with a synchronous motor, one need only 
ascertain that it is operating at synchronous speed. If the power 
source is a governor-controlled motor of adequate power, the problem 
is still relatively simple. If a small series- wound motor is used, it is 
only necessary to make sure that the rheostat adjustment will com- 
pensate for variations in line voltage. If, however, the source of power 
is a flat spiral spring or some similar energy-storage device, it is de- 
sirable to be able to study the speed of operation through the entire 
cycle during which the spring runs down. At the same time this 
equipment usually has little surplus power for the actuation of timing 
devices. Several methods have been used in the past, each of which 
has limitations. 

Mechanical-revolution counters are relatively simple and absorb 
little power but the results must be interpreted by computation and 
represent only the average speed during a given period of time. 

Mechanical or electromechanical direct-reading tachometers will 
give a continuous indication of speed but there must be available 
some moving part of the mechanism to which they can be attached 
and they invariably consume an appreciable amount of power in re- 
lation to the amount which is available. 

A sectored stroboscopic disk can also be used in one of two ways. 
This may be rotated by some part of the mechanism and observed by 
a pulsating light of known frequency. The disk may, alternatively, 
be rotated at a definite constant speed and illuminated by a light beam 
that is chopped or otherwise controlled by the shutter or some other 
part of the mechanism under observation. Special precautions must 

* Presented Apr. 23, 1947, at the SMPE Convention in Chicago. 
** Eastman Kodak Company, Rochester, N. Y. 




Vol 49, No. 5 

be taken to secure very brief light pulses in these cases or the sector 
pattern appears blurred. A high intensity of ambient light seriously 
decreases the readability of these devices. 

In either of these last two methods one sectored disk will show 
whether the mechanism is faster than, equal to, or slower than the 
speed for which the disk is calibrated. Two patterns are satisfactory 
to indicate whether the speed is at or within definite limits, but will 
not show the exact speed of the mechanism. 

The sectored disk or some part of the mechanism itself can be ob- 
served when illuminated by a pulsating light of variable but calibrated 

FIG. 1. Cine-camera speed indicator with camera in place for testing. 

frequency but this requires adept manipulation if the , speed changes 
as the motor runs down. 

One method which produces a permanent record is to photograph 
with the camera itself a pendulum or some other constant-frequency 
movement. This requires the processing and analysis of a length of 
film before the result is known. 

A satisfactory instrument to calibrate such apparatus should im- 
pose no load on the mechanism, give a direct, continuous indication of 
the speed, cover all speeds which may be encountered (8 to 64 frames 
per second), and have a linear scale. 

Fig. 1 is a photograph of a device which has been built to meet these 
requirements and does so quite closely. It is a rectangular cabinet 10 

Nov. 1947 



inches high, 15 inches wide, and 8 inches deep. At the left side of 
the face is a one-inch diameter hole against which the apparatus which 
is being checked is placed. At the right is a meter with a multiple 
scale, calibrated in frames per second. Along the bottom are a tell- 
tale light a range-selector switch, a push button for calibrating at 60 
cycles per second and an on-off switch. Adapters are provided to 
accommodate the various cameras. 

Fig. 2 is a photograph from the rear of the panel with the case re- 
moved. This shows the major electronic and optical elements. The 

FIG. 2. Rear view of chassis. 

potentiometer knob directly behind the meter is used in adjusting 
the calibration. 

Fig. 3 is a block diagram of the optical and the electronic systems. 
Light from a 6- to 8- volt, 15-candle-power lamp S passes through a 
half -silvered plate M which is set at an angle of 45 degrees. The trans- 
mitted light is converged by a simple condenser lens L into the aper- 
ture of the apparatus which is being checked. If the film-pressure 
plate in the camera is not bright or if the camera is being checked with 
film, a bright reflecting surface must be provided in the aperture. 
As the camera shutter opens the light is reflected back through the 



Vol 49, No. 5 

condenser L to the half -silvered plate M and a portion of this light is 
reflected to the cathode of a phototube C. As the shutter closes, no 
light reaches the phototube. 

The output of the phototube is fed to a 6SJ7 pentode whose internal 
plate resistance serves as the grid return of one half of an Eccles- 



J-LTL,, -LJL_ 




if r 






FIG. 3. 

Block diagram of cine-camera speed indicator. 

Jordan trigger circuit using a 6SN7 twin triode. One unit conducts 
when the shutter is open, the other when the shutter is closed. The 
plate voltage of each unit is a series of rectangular pulses of constant 
amplitude, one positive when the other is negative. The only variable 
is the frequency which is that of the shutter. 


I .2 



FIG. 4. Frequency-measuring circuit. 

These rectangular pulses which vary in frequency with that of the 
camera which is being calibrated are changed by resistance-capaci- 
tance differentiating networks into spiked trigger pulses which are 
fed into a frequency-measuring circuit. The output of the frequency- 
measuring circuit is a voltage which is dependent on input frequency 
and is read on the calibrated scale on the face of the instrument. 

Nov. 1947 



The most interesting part of the entire unit is the frequency-meas- 
uring circuit which is shown in Fig. 4. This includes two 2050 thyra- 
trons which act as a single-pole double-throw switch, shown symboli- 
cally at 1, 2. 

This switch is closed in one direction or the other as controlled by 
the two sets of trigger pulses. One of these provides a positive trigger 
pulse to one thyratron when the shutter opens, the other does the 
same to the second thyratron when the shutter closes. The negative 
pulses may be ignored as their magnitude is less than the ionization 
potential of the thyratrons they control. 

FIG. 5. Schematic diagram of cine-camera speed indicator. 

The first thyratron, fired by the shutter-opening pulse, charges C\ 
in its cathode circuit through a current-limiting resistor r to a fixed 
voltage equal to the regulated supply voltage less the tube drop. 
When the cathode reaches this potential, the negative grid regains 
control. The time constant rC\ is so small that the thyratron is always 
deionized before the shutter-closing trigger pulse arrives. This pulse 
fires the second thyratron and throws the switch to position 2, which 
discharges C\ into the much larger capacitor Cz shunted by a resistor 
R. This is repeated each time the shutter opens and closes. 

476 OWLETT Vol 49. No. 5 

The average current through R is i = fC\(E 2d ei), in which/ 
is the shutter speed in frames per second, E is the supply voltage, d is 
the voltage drop in each thyratron, and e\ is the peak voltage devel- 
oped across C 2 as C\ discharges. The average voltage e = iR is meas- 
ured by a vacuum-tube voltmeter calibrated in frames per second. 

If the time constant RC 2 of the output circuit is reduced so that C 2 
is always discharged between pulses, e\ is reduced to a small constant 
dependent on the ratio of the two capacitances. The calibration is 
then linear with frequency and the response to a change in frequency 
is immediate. The voltage would be in the form of a series of pulses 
with a sharp rise and an exponential decay. At the slower shutter 
speeds, a meter with normal damping would tend to follow these 
pulses rather than average them. 

As the RC 2 product is increased to smooth out the voltage across C 2 
and reduce the needle vibration of the indicating meter, e\ increases 
with frequency and a drooping characteristic is given to the calibra- 
tion curve. A lag in response is also introduced which increases with 
the RC>2 product. A detailed analysis is given at the end of the paper. 

The circuit for the instrument is shown in Fig. 5. Since the slowest 
shutter speed in general use is 8 frames per second, a compromise was 
adopted which gave a reasonably steady deflection at 5 frames per 
second without unduly slowing down the response of the meter. With 
a suddenly applied signal, the time required to reach 99 per cent of the 
final deflection is about 1.5 seconds. On higher ranges it would be ad- 
vantageous to increase the speed of response. This could be done in 
the range switching by progressively decreasing R. 

In later models a thermal time-delay switch will be incorporated to 
supply plate voltage to the thyratrons after all cathodes have warmed 


In this analysis, the two thyratrons are replaced by a single-pole double-throw 
switch thus neglecting the voltage drop in the tubes. It is also assumed that 
when the switch is closed in position 2, the voltage across the capa'citors reaches an 
equilibrium instantaneously. This is justified at the frequencies in which we are 
interested since the time constant is of the order of a few microseconds. 

Assume for the moment that the resistor R shunting Cz is removed, that C% is 
charged to a voltage CA and Ci to a voltage E. The charge on d is ECi, that on 


If the switch is now closed in position 2, the capacitors are in parallel 
and the charges are redistributed so that the voltage across C z is now 

Ed + e A Cz EC, e A C 2 C l Ci C 2 

Ci + C 2 Ci + C 2 "*" Ci + C 2 d + C 2 ~^ A Ci 

Ci C 2 

If we let - = K and = N, this reduces to 

e = K(E + MM). (1) 

If resistor R is now replaced, the capacitor will begin to discharge through 
it so that after any time t, the voltage will be e = K(E + NeA)e~ t/fRC z. Since the 
discharge period will always be the time to complete one cycle, t = I//, so 

e = K(E + NeA)*- 1 /'*. (2) 

Equations (1) and (2) will enable us to determine the transient and steady-state 
response of the circuit to any frequency. 

Initially, C\ is charged to a voltage E and C-z has no charge. The switch starts 
vibrating at a frequency /. As it closes in position 2 the first time, the voltage 
across C 2 becomes e = K(E + NO) = KE, and the switch returns to position 1 
to recharge C\ to a voltage E. After a time t = I//, the voltage across Cz will 
have dropped to e = K(E + NO}t~ l / fRC<i and the switch is ready to close in posi- 
tion 2 for the second time to add another increment of charge. 

The value of e over several cycles will be derived. Column A gives the value 
of e after the 1st, 2nd, ........ closure of the switch in position 2. Column B 

gives the value to whic